FIRE and MUD Contents

Petrology and Geochemistry of the 1991 Eruption Products of Mount Pinatubo

By Alain Bernard,¹ Ulrich Knittel,² Bernd Weber,² Dominique Weis,¹ Achim Albrecht,³ Keiko Hattori,⁴ Jeffrey Klein,⁵ and Dietmar Oles²

¹ Department of Geology, CP 160/02, Université Libre de Bruxelles, 50 Ave. F.D. Roosevelt, B-1050 Brussels, Belgium.

² Institut für Mineralogie und Lagerstättenlehre, Wüllnerstrasse 2, D-52056 Aachen, Germany. Now at: Institut für Geowissenschaften/Mineralogie Saarstr. 21, D-55099 Mainz, Germany.

³ Department of Chemistry, Rutgers University, Piscataway, NJ 08855; now at: EAWAG/ETH, Umweltphysik, 8600 Duebendorf, Switzerland.

⁴ Department of Geology, University of Ottawa, Ottawa, K1N 6N5 Canada.

⁵ Department of Physics, The University of Pennsylvania, Philadelphia, PA 19104.

ABSTRACT

The June 1991 activity of Mount Pinatubo erupted about 5 cubic kilometers of magma, mainly during the paroxysmal June 15 eruption. Most of the erupted magma was dacite with an overall constant composition (64.5 weight percent SiO₂) but with variable amounts of phenocrysts. Small volumes of andesite (56 weight percent SiO₂) characterized the June 7-12 dome-building phase and the June 12 eruption. This andesite contains inclusions of a primitive basalt (50.6 weight percent SiO₂). The dacite is extraordinarily enriched in sulfur (1,474 to 2,211 ppm sulfur), most of which is present as phenocrysts and microphenocrysts of anhydrite. The dacite also contains xenocrystic olivine, which, because of abundant chromite inclusions, is considered to be of juvenile nature. Major and trace element compositions of dacite and andesite display the same geochemical features as the other silicic rocks from the Bataan arc volcanoes of the Philippines. Strontium and neodymium isotopic compositions of all 1991 dacite and andesite and their phenocrysts are identical, with ⁸⁷Sr/⁸⁶Sr = 0.7042 to 0.7043 and ¹⁴³Nd/¹⁴⁴Nd = 0.51286 to 0.51298. The basalt contains slightly more radiogenic strontium and slightly less radiogenic neodymium. The isotopic signatures are in agreement with the volcano's location, which is relatively far from the North Palawan continental terrane (NPCT) collision zone as compared with the Macolod corridor, Luzon, or the Mindoro segment. These compositions suggest also that an unusually large contamination by marine sediments is not likely to explain the large sulfur enrichment observed in the Pinatubo magma. Dacite and andesite also have identical lead isotopic compositions characterized by high ²⁰⁸Pb/²⁰⁴Pb and ²⁰⁷Pb/²⁰⁴Pb for relatively low ²⁰⁶Pb/²⁰⁴Pb, while the basalt contains slightly more radiogenic lead. The basalt is likely to have been generated by partial melting of peridotite because of its high MgO content and high forsterite content of olivine pheno-crysts and the abundance of chromite inclusions. It is close to the primitive basalts found in the Macolod corridor.

Note to readers: Figures and tables open in separate windows. To return to the text, close the figure or table's window or bring the text window to the front.

INTRODUCTION

The 1991 eruption of Mount Pinatubo in the Philippines ranks among one of the largest volcanic eruptions of this century, with about 5 km³ of magma erupted (dense rock equivalent, DRE). The extraordinary feature of this eruption lies in the very large amount of SO₂ (20 Mt) injected into the stratosphere, well above the 7 Mt of SO₂ produced by the eruption of El Chichón volcano in 1982 (Bluth and others, 1992). Shortly after the eruption, the volcanic cloud produced by the conversion of this SO₂ into sulfuric acid aerosols spread rapidly in both hemispheres and significantly reduced the amount of sunlight reaching the Earth's surface (McCormick and Veiga, 1992; Stowe and others, 1992). Mean atmospheric and sea surface temperatures measured by satellites have now demonstrated that the cooling effect culminated in 1992 with a maximum temperature decrease of 0.6°C, which temporarily counteracted the effect of the greenhouse warming (Yan and others, 1992; Kerr, 1993). Due to enhanced concentrations of volcanic material at high altitude, a significant depletion in stratospheric ozone above Antarctica also occurred in 1991 (Hofman and others, 1992).

The high sulfur content of the Pinatubo magma is manifested by the presence of anhydrite phenocrysts in the pumices (Bernard and others, 1991; Fournelle, 1991; Knittel and others, 1991). Primary anhydrite is exceptional and has been reported only for very few young volcanic rocks (for example, Mount Lamington, Arculus and others, 1983; El Chichón, Luhr and others, 1984). Commonly, arc volcanics contain less than 200 ppm sulfur (S) (Gill, 1981) and typically contain less than 40 ppm S (Ueda and Sakai, 1984), while those from Pinatubo contain 1,500 to 2,400 ppm S (Bernard and others, 1991, Pallister and others, 1992).

A number of questions arise from the observation of this unusual eruption. The first question is what is the frequency of these sulfur-rich eruptions in the geological record? Up to now, the answers are only tentative:

1. Other silicic volcanoes may also be enriched in sulfur, but this is not known because most of them have never been studied in detail. Furthermore, anhydrite may have been overlooked or misinterpreted (as a product of hydrothermal alteration), or anhydrite that was originally present may have been leached away because this mineral is not very stable in tropical environments (Arculus and others, 1983; Luhr and others, 1984). Another problem is sample preparation. We have noticed that water-based thin-section preparation can remove anhydrite.
2. Sulfur may have been distributed unevenly in the subducted lithosphere and hence in the magma sources. The enrichment in sulfur may have an accidental origin, such as by contamination from metalliferous sediments. These sediments are usually only of limited extension and may have contributed only to the source region of a specific volcano.
   Another question is what is the likely source for the sulfur emitted into the atmosphere? Contrasting with the widespread occurrence of anhydrite, glass inclusions trapped in phenocrysts of dacite at Pinatubo show no evidence of sulfur enrichment. These inclusions contain only 60-90 ppm S and reveal that the preeruptive sulfur content dissolved in the silicate melt was low and close to the solubility limit for sulfur under the conditions of the magma reservoir (Westrich and Gerlach, 1992). The exsolution of the bulk sulfur dissolved in the melt during the eruption clearly cannot account for the 20 Mt of SO₂ produced. The alternative sources proposed for this excess sulfur are (1) breakdown of anhydrite during the eruption as was first suggested by Sigurdsson and others (1985) for the El Chichón eruption (Rutherford and Devine, 1991), or (2) presence of sulfur as a separate discrete phase before the eruption (Luhr, 1990; Westrich and Gerlach, 1992).
   Other aspects of interest concern the origin of the huge volume of relatively homogeneous dacite and its relationship to basalt, thought to have triggered the eruption (Pallister and others, 1992; this volume).
   We present here the results of a geochemical investigation carried out on the 1991 Pinatubo products. The main purpose of this investigation was to find geochemical evidence to explain the origin of this sulfur-rich magma in order to establish whether it is possible to identify ancient eruptions of anhydrite-bearing magmas on the basis of geochemical or petrological features.

GEOLOGIC SETTING

Mount Pinatubo, located on Luzon Island at long 120°22' E., lat 15°08' N., is the northernmost of several large stratovolcanoes forming the volcanic front along the Bataan segment of the Taiwan-Luzon arc (Defant and others, 1989) (figs. 1, 2). Volcanism of this arc is related to the subduction of the oceanic crust of the South China Sea along the Manila trench. Presently the arc is situated approximately above the 100-km depth contour line of the Wadati-Benioff zone (Cardwell and others, 1980; de Boer and others, 1980). West of Mount Pinatubo, the Zambales Ophiolite Complex is traversed by the west-northwest-trending Iba fracture zone, one of two major fracture zones that cut the ophiolite into three blocks (fig. 2). Thus, Mount Pinatubo appears to be located on the intersection of the Iba fracture zone and the north-south Bataan lineament (Wolfe and Self, 1983).

Figure 1. Tectonic setting of the Bataan segment of the Taiwan-Luzon arc.

Figure 2. The geology of the Bataan segment of the Taiwan-Luzon arc. The Zambales Ophiolite Complex is after Evans and others (1991). Offshore, the Scarborough seamounts mark the trace of the extinct spreading center of the South China Sea. Bathymetric contour interval 1,000 m. IG, Iba graben; SAG, San Antonio graben.

With an altitude of 1,745 m prior to the 1991 eruption, Mount Pinatubo was the highest volcanic edifice of this volcano chain. The volcano consists of an older, andesitic and dacitic stratovolcano and a younger, dominantly dacitic dome complex (Newhall and others, this volume). The last eruptions prior to 1991 produced voluminous ash-flow tuffs extending up to 20 km from the dome complex (Wolfe and Self, 1983). There are no records of historic eruptions prior to 1991 (since the 16th century); however, roaring vents near the summit and deposition of sulfur have been reported repeatedly. Dates obtained by the ¹⁴C method indicate that the last volcanic activity prior to the 1991 eruption occurred about 500 years ago (Newhall and others, this volume).

THE 1991 ERUPTION

Increased solfataric activity in August 1990 (PHIVOLCS, 1990), possibly in response to the July 16, 1990, earthquake along the Philippine-Digdig fault, was the first indication for a reawaking of Mount Pinatubo after about 500 years of dormancy. The eruptive activity started on April 2, 1991, with small explosions that created six new vents on the northern slope of the main cone. Starting May 16, small amounts of ash were ejected. This activity intensified on May 28 and reached a first culmination by the extrusion of a dome on June 7. Powerful eruptions producing andesitic ash started in the morning of June 12. The eruption reached its climax on June 15, when, within 12 h, an estimated 5 km³ of dacitic magma was erupted (W.E. Scott and others, this volume). Moderate to mild ash ejection continued until the end of August. (The account of the eruption is summarized from Pinatubo Volcano Observatory Team, 1991, and Sabit, 1992).

SAMPLES

Samples Pt6 through Pt32 were collected by Oles within 2 weeks of the June 15th eruption and samples Pin1 through Pin3 as well as Pin12 and Pin14 were collected during the same period by Bernard, who also collected samples Pin4 through Pin10 in September 1991. Additional samples were donated by Mylene Martinez (PHIVOLCS, samples Pt40, Pt41) and Chris Newhall (Pt51) and collected by Weber (Pt48, Pt52) in June 1992. Hattori collected samples of andesite with basalt inclusions in August 1992 (Pt50, Pt53). As pointed out by Newhall (oral commun., 1993), we cannot be absolutely sure that Pt50 and Pt53 are from the 1991 eruption; however, the chemical and mineralogical similarity of these last two samples with those analyzed by Pallister and others (this volume), together with their freshness, strongly supports this assumption.

ANALYTICAL PROCEDURES

XRF ANALYSES

Major elements were determined at Aachen University (samples with prefix "Pt") on fused discs with sample/flux ratio = 1:10. Trace elements were determined on undiluted pressed powder pellets. Samples with prefix "Pin" were analyzed at Liège University (Belgium) with the same analytical procedure.

STRONTIUM, NEODYMIUM, AND LEAD ISOTOPIC ANALYSIS

Most strontium (Sr) and neodymium (Nd) isotopic determinations and all lead (Pb) isotopic analyses were carried out at the Belgian Centre for Geochronology, Université Libre de Bruxelles, according to the following procedures.

Sample preparation.--About 3 to 4 g of rock powder were leached in hot, <100°C water in ultrasonic baths in order to dissolve the anhydrite. The rock was weighed before and after the H₂O bath in order to estimate how much anhydrite had been extracted. No precipitate was observed in any of the three solutions to which we added HCl 6 N in order to dissolve the anhydrite and to prevent gypsum precipitation. The rock powders were then leached for half an hour in warm 2.5 N HCl to eliminate secondary phases, if present, and to make sure that we would be left only with the primary silicates and glass. An additional leaching procedure (warm HCl 6 N) was applied to sample Pin4 to check the efficiency of the leaching procedure (see table 9). No difference was observed in any of the isotopic compositions analyzed, and samples Pin4 and Pin4a overlapped entirely within error limits. Again the powders were dried on the hotplate and weighed repeatedly until stable weight was obtained. About 100 to 200 mg of sample powder were dissolved in a mixture 6:1:1 HF-HNO₃-HClO₄ in a Teflon vessel.

Chemical separation.--Strontium and rare earth elements (REE) were separated with 2.5 N HCl and 4 N HCl, respectively, on a cation exchange column (Dowex 50Wx8). Neodymium was separated with 0.3 N HCl on a column prepared following Richard and others (1976). The procedure is described in detail by Weis and others (1987). Both columns were calibrated with ¹³⁹Ce. Lead was separated from the same starting sample solution by anion exchange columns with HBr-HCl following a method adapted from Manhès and others (1978). Lead and uranium (U) concentrations were measured by isotope dilution on the same solution (split before loading on columns and spiked with a mixed ²³⁵U-²⁰⁶Pb spike). Uranium was separated in an HNO₃ medium. All these operations were done in an overpressurized (>5 mm mercury) ultraclean laboratory and in laminar-flow cabinets. Blanks for the whole chemical procedure were <1 ng for Pb and <3 ng for Nd and Sr.

Mass spectrometry.- All the isotopic compositions, Sr, Nd, and Pb, were measured on the VG Sector54 7 collectors in dynamic mode for Sr (single tantalum (Ta) filaments) and Nd (triple, rhenium (Re) center and Ta side filaments) and in static mode for Pb (single Re filaments). The 2 errors on the mean for the isotopic compositions are less than 3x10^-5 for Nd and Sr isotopic ratios (that is, less than 0.01) and less than 0.1 for ²⁰⁶Pb/²⁰⁴Pb and ²⁰⁷Pb/²⁰⁴Pb, and less than 0.15 for ²⁰⁸Pb/²⁰⁴Pb. The Nd and Sr isotopic ratios were corrected for mass fractionation by using ¹⁴⁶Nd/¹⁴⁴Nd = 0.7219 and ⁸⁶Sr/⁸⁸Sr = 0.1194. Analyses of the NBS987 Sr standard gave ⁸⁷Sr/⁸⁶Sr = 0.710230+-5 (2_m on 67 measurements) and for the nNdb solution (Wasserburg and others, 1981), 0.51190+-2 (2_m on 15 measurements). Lead isotope results were corrected for mass discrimination (1.3+-0.3 per mass unit, Catanzaro and others, 1968) by repeated analyses of NBS 981.

Lead concentrations were measured with a Finnigan MAT-260 single collector mass spectrometer with double Re filaments. Rb and Sr concentrations were measured by X-ray fluorescence analysis to a precision of +-2 percent of values above 15 ppm.

Additional Sr and Nd isotopic data were collected by M. Feigenson (Rutgers University) and K. Hattori (University of Ottawa). At Rutgers, the measured ⁸⁷Sr/⁸⁶Sr ratio for NBS SRM 987 was 0.710248. ¹⁴³Nd/¹⁴⁴Nd was normalized to ¹⁴⁶Nd/¹⁴⁴Nd = 0.7219, and a value of ¹⁴³Nd/¹⁴⁴Nd = 0.511852 was obtained for the La Jolla standard. At Ottawa, isotopic determinations were made using Re double filaments for Sr and Nd isotopes in a Finnigan MAT-261 multicollector solid-source mass spectrometer. The Eimer and Amend standard gave ⁸⁷Sr/⁸⁶Sr = 0.70802 normalized to ⁸⁶Sr/⁸⁸Sr = 0.1194. ¹⁴³Nd/¹⁴⁴Nd for the La Jolla standard was 0.51186 normalized to ¹⁴⁶Nd/¹⁴⁴Nd = 0.7219.

BERYLLIUM ISOTOPES

The procedures for beryllium (Be) separation outlined here are based on methods developed for meteorite work. Adjustments were necessary because of the larger sample masses required (5 g) and the high aluminum (Al) concentrations. After addition of 1.5 mg ⁹Be carrier, open acid digestion using HF, HClO₄, HNO₃, and anion exchange to remove iron (Fe), the sample solutions were introduced into a large cation exchange column (allowing primary separation of Be from Al). After a buffered precipitation of Be and Al, still present as hydroxides, a second cation exchange procedure using small columns was performed to purify Be, which again was precipitated as hydroxide. This hydroxide was converted to BeO, which was loaded into the ion source of the Tandem Accelerator of the University of Pennsylvania, where the ¹⁰Be/⁹Be ratio was determined (Klein and Middleton, 1984). A blank measurement for the outlined procedure gave 6x10^-15 for the ¹⁰Be/⁹Be ratio, a relative error of 7 percent (see table 10).

PETROGRAPHY

The June 7-12 dome-building phase and the June 12 eruptions were characterized by the emission of andesite, which is considered to have been generated by mixing of basalt and dacite (Pallister and others, 1992; this volume). The basalt is represented by basaltic inclusions in the andesite.

The paroxysmal eruptions of June 15 produced two types of pumice: (1) white, phenocryst-rich pumice with large vesicles (>1 mm) and (2) gray, phenocryst-poor pumice with small vesicles (<1 mm). Both pumice types have identical chemical compositions (Pallister and others, 1992; table 1).

Table 1. Major and trace element compositions of the 1991 Mount Pinatubo eruptive rocks.

Rutherford (1993) and Rutherford and Devine (this volume) estimate that the preeruption pressure of the dacite at Pinatubo was 2.2 kbar with P_H2O=1.78 kbar. The oxygen fugacity is estimated to have been about 3 log units above the Ni-NiO (NNO) buffer (similar values are also calculated by Imai and others, this volume). The presence of cummingtonite rims around amphibole shows that the temperature was less than 810°C (Geschwind and Rutherford, 1992) and is estimated at 780°C.

DACITES

Phenocryst-rich dacitic pumice.-- The phenocryst-rich dacitic clasts are strongly porphyritic and vesicular pyroclastic rocks composed of euhedral and fragmented or fractured phenocrysts set in a vesicular, glassy groundmass that contains single glass shards and, to a lesser extent, fragments of the phenocryst assemblage. The pumiceous clasts contain 63 vol% vesicles, 16 vol% phenocrysts and 21 vol% glassy matrix (table 2). These data are compatible with the density of the pumiceous clasts (table 2) and close to the values reported by Pallister and others (this volume).

Table 2. Modal abundances of phenocryst-rich and phenocryst-poor dacites.

The phenocryst assemblage is dominated by plagioclase (29 vol%, on a vesicle-free basis), green hornblende (12 vol%, sometimes with cummingtonite rims), iron-titanium (Fe-Ti) oxides (0.8 vol%), and quartz (<0.1 vol%). Almost all samples contain microphenocrysts of anhydrite (0.1 vol%). These frequently contain inclusions of glass (often with a bubble) and apatite or are intergrown with apatite (fig. 3A). Biotite and clinopyroxene are present only in some of the analyzed clasts. Some samples contain glomerocrysts of plagioclase or polymineralic clots of quartz and plagioclase.

Figure 3. A, Anhydrite with glass (with bubble) and apatite inclusion. Anhydrite is about 0.43 mm in diameter. Plane polarized light. B, Xenomorphic olivine with chromite inclusions (black) in type 1 dacite. Olivine grain is about 0.35 mm across. Plane polarized light. C, Quartz xenocryst with corona of clinopyroxene, in basalt sample Pt53. The quartz grain has a diameter of 1.2 mm. Crossed polars. D, Contact between basalt (coarse grained, light colored) and andesite (fine grained, dark matrix). Note: In the basalt, the matrix between the plagioclase crystals (white) and amphiboles (dark gray) is filled by glass (light gray) and vesicles (white). Width of the photograph is 1.33 mm.

An interesting aspect of the dacites that erupted in 1991 is the occurrence of xenocrystic olivine (Reyes, Philippine National Oil Co., unpub. report, 1991; Knittel and others, 1991), the primitive nature of which is indicated by abundant chromite inclusions. These xenocrysts are commonly rimmed by brown amphibole. Occasionally the overgrowths also contain orthopyroxene. A few samples contain xenomorphic olivine grains without amphibole rims (fig. 3B).

The cryptocrystalline, highly vesiculated groundmass is dominated by single glass shards and light-colored matrix glass that encloses microphenocrysts and fragments of the phenocryst assemblage. Some clasts show a distinct lamination of the vesiculated groundmass. Sulfides were never observed in the groundmass. Very few micronic inclusions of pyrrhotite or chalcopyrite were observed in few plagioclase, amphibole, and magnetite phenocrysts (see detailed study of Hattori, this volume).

Phenocryst-poor dacitic pumice.--Although pheno-cryst-poor and phenocryst-rich dacitic pumices show similar chemical composition, there are various differences concerning mineralogical and textural appearance. The phenocryst content is much lower (10 vol%) and, on a vesicle-free basis, includes plagioclase (19 vol%), hornblende (4 vol%), and Fe-Ti oxides (0.4 vol%) (table 2). Other phenocrysts are rare. Only a few of the clasts studied contain quartz (<0.1 vol%), anhydrite (<0.1 vol%), biotite, and xenocrystic olivine.

Most phenocrysts in phenocryst-poor dacites are fragments of plagioclase and hornblende, much smaller than phenocrysts of the phenocryst-rich dacites. Euhedral crystals are significantly less abundant in phenocryst-poor dacite than they are in the phenocryst-rich dacite, and fragmentation is clearly evident from zoned plagioclase crystals in which the core is located at the grain edge. Therefore, we do not concur with the initial suggestion of Pallister and others (1992), who interpreted subhedral plagioclase to result from resorption. Note that Pallister and others (this volume) now consider fragmentation as the primary factor in generation of phenocryst-poor dacite.

The cryptocrystalline groundmass is dominated by highly vesiculated matrix glass that encloses microphenocrysts and fragments of the phenocryst assemblage. The average size of vesicles in phenocryst-poor pumice is smaller and results in a much denser appearance than that of the phenocryst-rich pumice. In addition, laminated texture of the groundmass is less common in the phenocryst-poor pumice than it is in the phenocryst-rich pumice.

DOME ANDESITE

The andesites are highly porphyritic rocks containing between 39 and 42 vol% phenocrysts, which are embedded in a fine-grained matrix (49-58 vol%). In contrast to the dacites, the andesites have significantly lower vesicle contents (3-7 vol%). This low vesicle content may not reflect the primary gas content of the andesite, as the dome andesite may have lost its volatile content by degassing before final emplacement.

The phenocryst assemblage comprises plagioclase, hornblende, clinopyroxene, olivine, and Fe-Ti oxides. Quartz xenocrysts and anhydrite are common but are not present in all samples. As in dacite, plagioclase is the dominant phenocryst (13-24 vol%) followed by hornblende (9-12 vol%), clinopyroxene (7-11 vol%), olivine (1-4 vol%), and Fe-Ti oxides (1-2 vol%). Quartz xenocrysts and anhydrite comprise up to 1.5 and 0.3 vol%, respectively, of some samples.

The fine-grained matrix is composed of clinopyroxene, plagioclase, and Fe-Ti microphenocrysts with abundant gray to light-brown matrix glass. Similar to the dacites, the andesitic groundmass contains fragments of phenocrysts.

BASALT

Basalt has been found as inclusions of variable size (millimeters to several centimeters) in andesite. The mineralogy is dominated by long, prismatic crystals of brown hornblende (30-35 vol%), which is very unusual for basaltic rocks. However, unlike in lamprophyres, the amphibole is not an early phase but postdates olivine and clinopyroxene.

Other phases are clinopyroxene (14-19 vol%), plagioclase (9-11 vol%), olivine (5-6 vol%), and Fe-Ti oxides and chromite (sum, 0.3-0.7 vol%). While clinopyroxene forms well-developed, euhedral crystals, the olivine is always marginally replaced by amphibole. Olivine usually contains tiny chromite inclusions. Pyrrhotite occurs as inclusions in the silicates (Hattori, this volume). The matrix of the mafic minerals is formed by small laths of plagioclase with brown glass and cavities in the interstices.

The basalt also contains several large crystals which show signs of instability and may be either xenocrysts or are derived from a small amount of evolved (dacitic?) magma mixed into the basalt. One of these phases is mottled, sieve-textured euhedral plagioclase. Some samples also contain quartz crystals rimmed by clinopyroxene coronas (fig. 3C). Commonly, relatively large glass patches are associated with the quartz xenocrysts. Furthermore, there are decomposed mafic minerals. Textural evidence suggests that they are amphiboles that have been replaced by clinopyroxene, and perhaps also clinopyroxenes.

The contacts between basalt and andesite are generally sharp (fig. 3D). Cuspate and interfingering contact relations between the two suggest that both rocks were partially molten when they came into contact.

MINERALOGY

Plagioclase.--The plagioclase crystals in dacite are relatively sodic, with rims often more calcic than cores (cores An_30-40 and rims up to An₄₅, fig. 4). In detail, the plagioclase phenocrysts show complex zoning (see Pallister and others, this volume). Plagioclase in evolved dacites of Mount St. Helens have similar compositions (Smith and Leeman, 1987), but those in dacites from northern Luzon are less albite rich and show much greater compositional variability within single samples (Knittel and others, unpub. data, 1992).

Figure 4. Orthoclase (Or) versus anorthite (An) content (percent) of plagioclase from dacite (filled symbols = core compositions, open symbols = rim compositions). Squares, phenocryst-poor dacites; circles, phenocryst-rich dacites.

Amphibole.--The green amphiboles in dacite are fairly homogeneous. With 6.4-6.9 Si (on the basis of 23 O) and around 1.6 Ca, they approach the composition of edenite (fig. 5, table 3). The amphibole in the andesite appears to be even slightly higher in Si (though this could be a sample bias; only one andesite was investigated). The high Si content is typical for amphibole in evolved rocks. The brown amphibole in basalt is distinctly less Si rich (Si~6) but contains more Ca (Ca~1.8).

Table 3. Microprobe analytical compositions of amphibole.

Figure 5. Silicon versus calcium content for amphiboles (number of atoms on the basis of 23 oxygen atoms). Diamonds, basalt; open squares, andesite; filled squares, dacite.

Clinopyroxene.-- The clinopyroxenes show little compositional variability. They are magnesian salites generally low in titanium (Ti) and sodium (Na) (table 4).

Table 4. Microprobe analytical compositions of pyroxene.

Anhydrite.-- Anhydrite contains significant SrO (~0.23 wt%, table 5), which is, however, only half the SrO content of El Chichón apatite (0.58 wt% SrO; Luhr and others, 1984).

Table 5. Microprobe analytical compositions of anhydrite.

Biotite.--Biotite is intergrown with amphibole in rims around olivine, in andesite sample Pin6. This biotite is characterized by extremely high but variable Mg/(Mg+Fe) of 0.74-0.89 (table 6). The highest values are probably due to the fact that this biotite replaces Mg-rich olivine without significant addition or loss of Fe and Mg. Mg/(Mg+Fe) of biotite crystals in dacite is about 0.67-0.70 (table 6; see also Imai and others, this volume). These values are still high for such evolved rocks but may reflect the high oxidation state, hence high Fe³⁺ and low Fe²⁺ contents of the melt (see Imai and others, this volume, for a more detailed discussion).

Table 6. Microprobe analytical compositions of biotite.

Olivine.-- Olivine is a major constituent of the basalts and andesites. Xenocrystic olivines are also found in several samples of dacitic pumice. Petrographic features of the olivines have already been described. Regardless of the host rock, olivine core compositions are uniformly Fo_86-88 and rims have Fo_85-87 (table 7). Pallister and others (this volume) and Imai and others (this volume) found relatively Fe-rich olivine in dacite (Fo_~81), but we found only the Fo-rich olivines in three dacite samples; we have at present no explanation for this discrepancy. Because the olivine grains commonly are replaced by amphibole along the margins and some grains in dacite appear to have experienced some kind of abrasion or resorption as indicated by their subhedral shape, the "original" rims may have been more Fe rich.

Table 7. Microprobe analytical compositions of olivine.

Chromite.--The olivine grains contain tiny, commonly euhedral chromite crystals, characterized by Cr/(Cr+Al) = 0.64-0.68, Mg/(Mg+Fe²⁺)= 0.33-0.48 (table 8) and Fe³⁺ contents that are high relative to chromites from midocean ridge basalt (MORB) (Dick and Bullen, 1984) and primitive basalts from the Macolod corridor, located to the south (fig. 6). The chromite inclusions in basalt and andesite have relatively constant Fe³⁺ contents, while dacites have slightly more oxidized chromites that may have reequilibrated. This is suggested by the observation that the most highly oxidized chromite (filled stars in fig. 6) occurs within a marginally oxidized olivine grain. Similarly high Fe³⁺ contents have been reported by Nye and Reid (1986) for chromite inclusions in olivine from primitive basalts from Okmok Island, Aleutians (fig. 6). Chromite pheno-crysts in basalt that are not included in olivine are strongly zoned, with Al/Mg-rich cores surrounded by magnetite rims.

Table 8. Microprobe analytical compositions of chromite inclusions in olivine.

Figure 6. Compositional variation of chromite inclusions in olivine in the different Pinatubo eruption products, plotted in the Cr-Al-Fe⁺³ triangle. Chromite inclusions in olivine from basalt are shown as filled squares (Pt42) and circles (Pt53). Those from andesite are shown by open squares and circles (two different parts of sample Pt42), and those from dacite as filled (Pt21) and open (Pt9) stars. Fields for Cr-spinels from Macolod basalts (Knittel and Oles, 1995), Taal basalt (U. Knittel, unpub. data, 1992), and basalts from Okmok Island (Nye and Reid, 1986) are shown for comparison; spinels from MORB typically plot below and to the left of the Macolod field (Dick and Bullen, 1984).

GEOCHEMISTRY

Clasts of both types of dacite have virtually identical compositions, with 64.5 wt% SiO₂ on average. Al₂O₃ is slightly higher in the phenocryst-poor dacite (16.7 versus 16.0 wt%), a fact that may reflect a sampling or analytical bias because all phenocryst-rich dacites analyzed in Liège fall at the low end of the range in Al₂O₃ observed for the samples analyzed in Aachen. A clear analytical bias exists for barium (Ba), where all Liège analyses show higher values than the Aachen analyses. Ba data given by Pallister and others (this volume) show better agreement with the Liège data. With regard to all other major and trace element abundances, the results obtained in the two laboratories are compatible within analytical uncertainty.

Sulfur content in the dacite ranges from 1,470 to 2,210 ppm. The only andesite analyzed for sulfur (Pin6) is slightly enriched in sulfur relative to the dacites. Two small clasts (5-10 cm in diameter) collected in September 1991 (Pin9 and Pin10, with 421 and 353 ppm S, respectively) clearly show the effect of leaching of anhydrite by rainwater during the first rainy season following the eruption.

Harker variation diagrams of selected major and trace elements of the 1991 Pinatubo rocks are shown in figures 7 and 8. In these figures, the Pinatubo compositions are compared to other volcanics from the Luzon arc reported by Defant (1985) and Defant and others (1991). The 1991 dacites and andesite compare well with the most silicic rocks of the Bataan arc-front volcanoes segment, but it should be noted that volcanics with SiO₂>62 wt% are quite rare in the Bataan segment. The dacites of Pinatubo display the same geochemical characteristics that Defant and others (1991) observed for the Bataan arc-front volcanoes segment: significantly lower K₂O values than the volcanics from the Bataan behind-arc volcanoes, Macolod, and Mindoro segments. In detail, however, the Pinatubo dacites are slightly enriched in MgO, Na₂O, and Sr and are depleted in Al₂O₃ and yttrium (Y).

Figure 7. Major elements plotted against SiO₂, for 1991 Pinatubo and other eruption products. Crosses, Pinatubo dacite, andesite, and basalt (together); diamond, Bataan behind-arc volcano lavas; square, Bataan arc-front volcano lavas; circle, Macolod corridor lavas.

Figure 8. Trace elements plotted against SiO₂, for 1991 Pinatubo and other eruption products. Symbols as in figure 7.

It is interesting to compare the Pinatubo dacite with that of Mount St. Helens, because many data are available for these (for example, Smith and Leeman, 1987). Compared to 1980 Mount St. Helens dacite, Pinatubo dacite is enriched in SiO₂ and lower in Al₂O₃ and FeO. These differences exist also if older Mount St. Helens dacites are considered, except for a few lavas of the Kalama period (Smith and Leeman, 1987). Large-ion lithophile element (LILE) abundances in Pinatubo dacite are slightly higher (in agreement with slightly higher K₂O), zirconium (Zr) abundances are identical, and niobium (Nb) abundances are lower than in dacite from Mount St. Helens.

ISOTOPES

STRONTIUM AND NEODYMIUM ISOTOPIC COMPOSITIONS

Sr and Nd isotopic compositions were measured on type 1 (phenocryst-rich) and type 2 (phenocryst-poor) dacites of Pinatubo. Both types give identical compositions within the error limits of the analytical method, even with some of the analyses carried out in different laboratories (table 9). These compositions are in close agreement with the data reported by Castillo and Punongbayan (this volume). A sample of the andesite found as an inclusion in a dacitic pumice clast (Pin6, table 1) and similar to the andesite scoria described by Pallister and others (1992) had Sr and Nd isotopic compositions identical to those of the dacites (table 9). In contrast, the basalt contains slightly more radiogenic Sr and slightly less radiogenic Nd than the dacites (table 9).

Table 9. Strontium, neodymium, and lead isotopic compositions of the 1991 Mount Pinatubo rocks and phenocrysts.

⁸⁷Sr/⁸⁶Sr compositions were also obtained for mineral separates (table 9). Because of the small amounts of anhydrite present in the samples, this mineral was extracted by leaching with distilled water. Since no other water-soluble mineral was observed in the dacite, and because the anhydrite is Sr rich (SrO up to 0.23 wt%), the analyses of the leachates are considered to be representative of the Sr isotopic compositions of this sulfate. In the four leachate samples analyzed, the anhydrite has the same Sr compositions as the whole rocks. This similarity supports the assumption of a primary magmatic origin for this sulfate, as suggested in the previous sections.

Data obtained from other phenocrysts, plagioclase, and amphibole show likewise that they have crystallized in isotopic equilibrium with the silicate melt.

⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd isotopic compositions of the 1991 Pinatubo products are at the high ¹⁴³Nd/¹⁴⁴Nd and low ⁸⁷Sr/⁸⁶Sr ends of the spectra observed for the Bataan segment (fig. 9). Relative to most other volcanoes of the Bataan segment, they show a very slight shift toward high ⁸⁷Sr/⁸⁶Sr (fig. 9). Another Bataan segment sample that shows such a shift is from Mount Natib (Defant and others, 1991), a nearby volcano that appears to be heterogeneous with respect to Sr isotopic composition (Knittel and Defant, 1988, reported ⁸⁷Sr/⁸⁶Sr ratios of 0.7042-0.7051 from Mount Natib).

Figure 9. ¹⁴³Nd/¹⁴⁴Nd versus ⁸⁷Sr/⁸⁶Sr for volcanic rocks from the southern part of the Taiwan-Luzon arc (North Luzon, Bataan, and Mindoro segments, and Macolod corridor) compared to Pinatubo samples. Inset shows all Pinatubo data in detail (this paper; Castillo and Punongbayan, this volume). Main fields for data from the North Luzon, Bataan, and Mindoro segments are indicated. Data sources: North Luzon-Defant and others (1990); Bataan and Mindoro segments-Knittel and others (1988), Defant and others (1991), Mukasa and others (1994); Macolod corridor-Knittel and others (1988), Knittel and others (1992).

Knittel and others (1988), Knittel and Defant (1988) and Defant and others (1990, 1991) have shown that recent volcanism along the Luzon volcanic arc shows systematic regional variation. In northern Luzon (North Luzon segment) the volcanics have Sr-Nd isotopic compositions similar to older (middle Tertiary) plutonic rocks and to recent volcanics associated with westward subduction along the Philippine trench (eastern coast of Luzon, fig. 1; Knittel-Weber and Knittel, 1990). Toward the south (Bataan and Mindoro segments), the volcanics have progressively less radiogenic Nd and more radiogenic Sr values. Knittel and others (1988) and Defant and others (1991) suggested that these differences can be attributed to the incorporation of more crustal material into the source region of these mantle-derived magmas. The likely source of the crustal material is subducted fragments from the North Palawan continental terrane (NPCT), which collided with the Philippine archipelago during the late Miocene, or sediments derived from the NPCT (fig. 1).

The 1991 Pinatubo dacites and their phenocrysts have ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd isotopic signatures in agreement with the volcano's location relatively far from the NPCT, as compared with the Macolod corridor or the Mindoro segment. Therefore, it is unlikely that the sulfur enrichment observed in the Pinatubo magma is resulting from unusually large contamination by marine sediments (evaporites).

LEAD ISOTOPIC COMPOSITION

Lead isotopic compositions were measured on two crystal-rich dacites (Pin4 and Pin7), an andesite (Pin6) and a basalt inclusion (Pt53). They are listed in table 9 and are presented graphically in figure 10. The dacite and the andesite have virtually the same composition, while the basalt contains slightly more radiogenic Pb.

Figure 10. Lead isotopic compositions of 1991 Pinatubo eruption products (this study; Castillo and Punongbayan, this volume) compared to other eruption products from the Taiwan-Luzon arc. Data sources: Batan Island-Vidal and others (1989), McDermott and other (1993); Taal volcano, precollision volcanics and postcollision rocks far from the collision zone-Mukasa and others (1987, 1994); sediments-McDermott and others (1993), Mukasa and others (1987). The observed steep trends indicate mixing between "precollision-type" mantle and subducted sediment.

Mukasa and others (1987) have shown that volcanic rocks from the Philippines extruded after the collision episode and close to the NPCT collision zone (that is, southern Luzon, Panay, and Mindoro) are characterized by high ²⁰⁸Pb/²⁰⁴Pb and ²⁰⁷Pb/²⁰⁴Pb and relatively low ²⁰⁶Pb/²⁰⁴Pb. Lead isotopic compositions obtained from various older rocks and rocks remote from the collision zone (the West Philippines volcanic arc) contain less radiogenic Pb (Mukasa and others, 1987). Despite the relatively unradiogenic Sr and radiogenic Nd, the 1991 Pinatubo rocks fall well into the field of recent volcanic arcs (postcollision) located close to the NPTC collision zone. Similar Pb isotopic ratios were also observed in lavas and mantle nodules from Batan Island, a volcanic island located north of Luzon (fig. 10), and interpreted to indicate the presence of lead derived from subducted sediments in the source region (Vidal and others, 1989).

¹⁰Be

The presence of significant amounts of subducted sediment potentially can be monitored by ¹⁰Be, a short-lived cosmogenic nuclide (half-life 1.5 m.y.) that is produced in the atmosphere, from where it is removed by rain or adsorbed by solid particles. ¹⁰Be concentrations in marine sediments are particularly high because of low sedimentation rates (on average 5,000x10⁶ atoms ¹⁰Be/g sample in young sediment, Brown, 1984). For ¹⁰Be to be found in arc magmas, subduction of the uppermost sediment layer and incorporation of that sediment into the magma source in <9 m.y. is required (Morris and others, 1990). Thus, lack of detectable ¹⁰Be does not imply the lack of a sediment component in the source region; ¹⁰Be could also be undetected because the uppermost sediment layer was scraped off or because subducted components were recycled too slowly (Woodhead and Fraser, 1985).

Two samples of phenocryst-rich pumice, Pt12 and Pt25, which were also analyzed for their Sr and Nd isotopic composition (table 9), were chosen for ¹⁰Be analysis. These samples contain 1.96 and 2.77x10⁶ atoms ¹⁰Be/g sample (table 10), respectively, which is higher than the value of 0.8x10⁶ obtained for Mount Mayon (Tera and others, 1986), the only other Philippine volcano for which ¹⁰Be data are available. These data are evidence that the Pinatubo dacites contain a sedimentary component that has been recycled through the subduction system.

Table 10. Beryllium isotopic compositions of whole-rock pumice.

SULFUR ISOTOPES

³⁴S for whole-rock pumice, water-soluble anhydrite, and insoluble sulfur were obtained on both types of dacites and are listed in table 11. These data show a relatively large spread in ³⁴S values, ranging from +6.9 to +12.2, and bracket ³⁴S values reported elsewhere for the dacite: +7.8 to + 9.0 (Imai and others, this volume). This range extends even to lower values (³⁴S/³²S = 6.3) for soluble sulfate measured on the andesite (Pin6) and suggests a relatively heterogeneous composition for sulfur. Individual anhydrite phenocrysts of June 12 products show a very broad range of ³⁴S values from 3 to 16, but phenocrysts of June 15 products show a narrower range of ³⁴S values (McKibben and others, this volume). McKibben and others conclude that contamination of the magma by extraneous sulfur at shallow levels seems necessary to explain the observations on June 12 samples. This conclusion is also made by Hattori (this volume) on the basis of a study of sulfides in the 1991 Pinatubo eruption products.

Table 11. Sulfur isotopic compositions of dacite, andesite, and ash samples from Mount Pinatubo.

Two fresh ash samples collected on July 4, 1991 (Pin12 and Pin14), were analyzed also. The sulfur content of these ash samples (table 1) is markedly higher than the sulfur content of the dacite because a significant amount of sulfate derived from the gas phase in the eruptive column is adsorbed on these samples. In consequence, the measured ³⁴S is a mixed contribution of about 30 to 35 percent of sulfate from the gas phase and 65 to 70 percent of sulfate derived from the anhydrite microphenocrysts. These values are within the range of data obtained on the dacite.

DISCUSSION

GENESIS OF THE BASALT

The basalt, found as inclusions in the June 7-12, 1991, andesite lava, is likely to have been generated by partial melting of peridotite because of its high Mg content, the high Fo content of olivine phenocrysts, and the abundance of chromite inclusions in olivine. Furthermore, typical olivine core compositions are in equilibrium with a melt having Mg/(Mg+Fe²⁺) = 69 (assuming a partition coefficient K_d Fe/Mg_{(olivine/liquid)} = 0.3; Roeder and Emslie, 1970), and this is the observed Mg/(Mg+Fe²⁺) of the basalt, assuming Fe²⁺/(Fe³⁺+Fe²⁺) = 0.85 (Nicholls and Whitford, 1976). This equilibrium between core compositions and the melt in turn implies that the basalt was not highly oxidized or was oxidized only at a late stage. This conclusion is supported by the presence of abundant sulfides in the basalt (Hattori, this volume).

The fact that the basalt appears to be in equilibrium with olivine Fo₈₈ would suggest that the magma is not far removed from its primary composition by fractional crystallization. However, the relatively low nickel (Ni) content of the whole rock (131 ppm Ni) and the olivines (0.25% NiO at Fo₈₈) suggests that the melt does not represent a primary composition but is slightly evolved.

Despite this limitation, it is tempting to plot the composition into the (Jd+CaTs)-Qz-Ol projection of the basalt tetrahedron of Falloon and others (1988) to obtain a rough estimate of the pressure at which the melt equilibrated with the residue. The position of the basalt in the plot suggests a value of about 14 kbar (fig. 11) and also suggests that the melt was in equilibrium with olivine+orthopyroxene+-spinel, that is, with a harzburgitic assemblage. Since the basalt has fractionated some olivine, the equilibrium pressure must have been slightly higher (addition of 10% olivine would give an equilibrium pressure of about 18 kbar; addition of more olivine does not seem to be justified in view of the high Mg# of the basalt). In particular, the effect of olivine fractionation may have been increased by some assimilation, which is indicated by the presence of rare quartz, amphibole, and large plagioclase crystals in the basalt.

Figure 11. Basalt from Pinatubo (large square) plotted into the (Jd+CaTs)-Qz-Ol (jadeite+Ca tschermakite - quartz - olivine) projection of the basalt tetrahedron of Falloon and others (1988). Shown are melt compositions derived from moderately depleted MORB-pyrolite (MP) and highly depleted Tinaquillo lherzolite (TQ). Subvertical lines are melts in equilibrium with olivine. For melts derived from TQ, compositions to the right of the dotted "cpx-out" line are in equilibrium with olivine+orthopyroxene, and those to the left with olivine+orthopyroxene+clinopyroxene. The small square represents Pinatubo basalt with 10 percent olivine added (see text). The arrows indicate the effect that fractionation and contamination have on compositions plotting in the same area as basalt of Mount Pinatubo. The composition of the basalt suggests earlier equilibration at about 14 kbar.

Primitive basalts are very rare in island arcs. Mg/(Mg+Fe²⁺) ratios of typical eruption products of arc-front volcanoes of the Bataan segment are 0.50 to 0.55 (Defant and others, 1991). However, primitive basalts have been erupted by numerous small eruption centers within the Macolod corridor in southwestern Luzon (Oles, 1988; Förster and others, 1990; Knittel and others, 1992). Olivine core compositions in these basalts reach Fo₈₉, and Mg/(Mg+Fe²⁺) ratios of the more primitive basalts are 64 to 73. Chromite inclusions have similar, though usually slightly lower, Cr/(Cr+Al) than those in Pinatubo olivine xenocrysts (fig. 6).

The basalt inclusions of Mount Pinatubo closely resemble the primitive Macolod basalts in their major element composition, except for their lower Al₂O₃ content (fig. 12), a feature they share with basalts erupted from Taal, located in the western part of the Macolod corridor (Miklius and others, 1991). However, Taal basalts are distinctly more CaO rich compared to the Pinatubo basalt.

Figure 12. Major element compositions of basalts (in weight percent) from Pinatubo (this paper; Pallister and others, this volume) compared to other basalts from the Philippines (Macolod basalts-Knittel and Oles, in press; basalts from Taal-Miklius and others, 1991; basalt from Malinao (southern Luzon, Bicol arc-Knittel-Weber and Knittel, 1990).

The scarcity of primitive basalts in arcs suggests that such magma may reach the surface in this environment only under special circumstances. The Macolod corridor (fig. 1) is a graben striking northeast-southwest that is roughly perpendicular to the volcanic arc (Voss, 1971; Oles, 1988; Förster and others, 1990). The presence of primitive basalt in the 1991 Pinatubo eruption products suggests that this volcano also may be located in an extensional setting. Indeed, Mount Pinatubo is situated at the intersection of the volcanic front and the Iba fracture zone, a major east-southeast-trending structure that cuts the Zambales Ophiolite Complex (fig. 2). Accordingly, we suggest that the magmatic activity of Mount Pinatubo is controlled by its position relative to the underlying plate and its location within what may be a tensional regime. The Iba fracture zone may be part of an incipient horst and graben system which, like the Macolod corridor, accommodates the differential movements resulting from eastward-directed subduction along the Manila trench to the north and westward-directed subduction along the Philippine trench to the south.

Additional evidence linking the basalt genesis to the setting of the Iba fracture zone is provided by the REE pattern of the basalt (data in Pallister and others, this volume). Pallister (written commun., 1993) noted that the Pinatubo basalt is highly enriched in the light REE (LREE) (La/Yb as high as 13), while other volcanic rocks along the volcanic front of the Bataan segment are only slightly enriched in LREE (La/Yb mostly about 4; Defant and others, 1991) (fig. 13). Similar high fractionation of LREE and heavy REE (HREE) is observed for some basalts of Mount Arayat, the only major stratovolcano located behind the volcanic front of the Bataan segment, 45 km east-northeast of Mount Pinatubo. Most Arayat basalts have La/Yb = 5 to 10; only three samples have La/Yb = 14 to 18; Defant and others, 1991; Bau and Knittel, 1993). Mount Arayat may owe its existence to some "weak" zone within the crust, though we note that it does not lie along strike of the Iba fracture zone.

Figure 13. Lanthanum/ytterbium versus ytterbium for basalt of Pinatubo (this paper, Pallister and others, this volume), basalts of Arayat (Bau and Knittel, 1993), and various volcanic rocks from the volcanic front of the Bataan segment (Defant and others, 1991).

GENESIS OF THE DACITE

There is no generally accepted model for the genesis of dacite at convergent plate boundaries, and possibly there is no single unique mechanism responsible for the generation of such rocks. Commonly, the following mechanisms for genesis of arc dacite are considered (after Reid and Cole, 1983).

1. Fractional crystallization of basic precursors.
2. Mixing of rhyolitic and more basic magmas (Graham and Worthington, 1988); alternatively, a basic magma may have been massively contaminated by silicic rocks.
3. Generation of dacitic melt by melting of basaltic rocks either in amphibolite facies at crustal levels (Giese and others, 1986; Smith and Leeman, 1987) or in eclogite facies within the subducted slab (Drummond and Defant, 1990; Defant and Drummond, 1993).

The relatively high concentrations of incompatible trace elements in the basalt (K, Rb, Ba, Zr; fig. 8) rule out that the dacites are derived by fractional crystallization from the basalt that is found as inclusions in the andesite. For the same reason, Smith and Leeman (1987) concluded that the Mount St. Helens dacite is not derived from basalts erupted from Mount St. Helens. In addition, the difference in isotopic composition between dacites and basalt rules out a cogenetic relation. Alternatively, there is evidence that some volcanoes produced different types of basalt (for example, at Mount St. Helens; Leeman and others, 1990) and, thus, the dacites of Pinatubo could be derived from a basaltic precursor that is as yet not discovered at Mount Pinatubo. However, REE data for the dacite provided by Pallister and others (this volume) and Castillo and Punongbayan (this volume) show that the dacite is depleted in HREE, and model calculations carried out by Martin (1987) show that such patterns probably cannot be generated by fractional crystallization of mantle-derived basalt.

In view of the suggestion of Pallister and others (1992) that the eruption was triggered by the intrusion of basaltic magma into a reservoir containing evolved melt, could magma mixing be a process for the generation of the dacite? According to Pallister and others (1992), dacite was one of the endmembers, and mixing produced andesite. Therefore, if the Pinatubo dacite had been produced by mixing, this must have occurred prior to the events that led to the 1991 eruption. In addition, the most evolved rocks known to have been erupted from Mount Pinatubo are dacites with 64 to 66 wt% SiO₂ (Defant, 1985; Pallister and others, this volume).

The Pinatubo dacites have some features pointing to an origin by melting of basaltic precursors, which typically are rich in Al₂O₃ (fig. 14) and are characterized by normative Ab/An >1 and low normative Or content (fig.15; Helz, 1976; Rapp and others, 1991). Surprisingly, the Pinatubo dacites have about the lowest Al₂O₃ contents of all dacites in Luzon (fig. 7 and Knittel and others, unpub. data, 1992), but many Archean trondhjemites, believed to have been generated by melting of basaltic protoliths, likewise have slightly lower Al₂O₃ than predicted by the experiments (Rapp and others, 1991).

Figure 14. Al₂O₃ versus SiO₂ plot comparing dacites from Pinatubo and experimentally produced melts in the range SiO₂=56-71 wt% (Helz, 1976; Rapp and others, 1991). Dacites from Mount St. Helens (Smith and Leeman, 1987) are plotted for comparison. Note that melts produced at 5 to 8 kbar have significantly higher Al₂O₃ contents than both Mount Pinatubo and Mount St. Helens dacites.

Figure 15. Dacites from Pinatubo and Mount St. Helens, plotted in the normative albite-orthoclase-anorthite triangle and compared to compositions produced by melting basalts experimentally. CIPW norms were calculated with FeO assumed to be 0.85FeO_total. Data sources: Kilauea basalt and Picture Gorge basalt from Helz, 1976; basalt numbers 1-4 from Rapp and others (1991).

Compared to lavas of other volcanoes from the volcanic front of the Bataan segment, the Pinatubo dacites are notable for their high Sr and low Y contents (fig. 16). In recent papers, Drummond and Defant (1990) and Defant and Drummond (1990) identified these features as characteristics of arc magmas that contain significant melt contributions from subducted oceanic crust that has been metamorphosed to eclogite.

Figure 16. Strontium/yttrium plotted against yttrium for dacites from Mount Pinatubo (this paper), dacites from northern Luzon (Knittel and others, in press), and andesites and dacites from Mariveles (Defant, 1985). Fields for typical calc-alkaline lavas, considered to be derived from basic precursors, and for melts thought to have been generated by melting of MORB in eclogite facies, are after Drummond and Defant (1990) and Defant and Drummond (1990).

In principle, the hypothesis that the Pinatubo dacite was generated by partial melting of eclogite in the subducting slab can be tested on the basis of the REE contents of these rocks, because garnet is the only petrologically significant phase able to fractionate the middle REE (MREE) from the HREE and to cause significant HREE depletion (see models calculated by Martin, 1987). The Pinatubo dacites show significant fractionation between LREE and HREE abundances (La_N/Yb_N = 8, where the subscript "_N" indicates a value that is normalized to chondritic values) and HREE depletion relative to "common" arc rocks (Yb_N = 6). We note, however, that the LREE/HREE fractionation and the HREE depletion of the Pinatubo dacite is smaller than in any of the rocks considered by Drummond and Defant (1990) to be partial melts derived from eclogite.

To further evaluate the origin of the dacites, we have calculated hypothetical REE patterns for the source of the Pinatubo dacite, assuming simple batch melting and that the residue consists only of garnet and clinopyroxene. A simple melting model seems to be justified, as small-volume melts probably cannot reach crustal levels because of the interaction with peridotitic wall rocks within the mantle.

A major problem of such calculations is choosing the distribution coefficients for REE in garnet and siliceous melts, because the range of values is wide (for example, Irving and Frey, 1978). Previously, low (Kay, 1978) as well as high (Martin, 1987) K_d(REE) have been used to calculate models supporting the eclogite melting hypothesis. Though the data given by Irving and Frey (1978) suggest that REE are significantly more compatible in garnets found in siliceous rocks, we have considered two extreme data sets. Calculations assuming low K_d(REE) values (those of Kay, 1978) require about 30% of garnet in the residue to give REE pattern with Yb_N around 10 (fig. 17), a value commonly observed for MORB including South China Sea basalts (which are represented by the East Taiwan Ophiolite according to Chung and Sun, 1992; REE data from Jahn, 1986). A higher garnet content in the residue would produce unrealistic REE patterns with La_N/Sm_N>1 and Sm_N/Yb_N<1 (that is a wavy REE pattern); lower garnet content in the residue produces Yb_N lower than in MORB. Calculations using high K_d(REE) constrain the amount of residual garnet to about 10%; otherwise, the hypothetical source would be highly HREE enriched. MORB-type REE patterns are obtained with approximately 10% garnet and 30% partial melting (fig. 17).

Figure 17. Calculated REE patterns of hypothetical sources of dacite from Pinatubo, when it is assumed that the melts were generated by melting of eclogite. Shaded fields are patterns for MORB with boundaries based on E- and N-MORB from the South China Sea (Jahn, 1986; Chung and Sun, 1992). Degree of partial melting and residue compositions are given in the explanation (gt=garnet, px=clinopyroxene, F=melt fraction). Chondritic values after Nakamura (1974). K_d, distribution coefficient.

To summarize, the source models calculated on the basis of the REE content of the Pinatubo dacite require either (1) a low garnet content of the residue, which is not supported by the experimental results of Rapp and others (1991), or (2) that low K_d(REE) apply for the genesis of Mount Pinatubo dacite, a condition that is not supported by what is known about the variation of K_d(REE) in garnet.

Alternatively the source could have HREE contents significantly different from those of MORB. Thus, while the calculations do not disprove the eclogite melting hypothesis, due to the uncertainties involved, they provide no evidence for such a model.

The low garnet content in the residue, suggested by the calculations, would, however, be compatible with the assumption that the dacites were generated by melting of garnet-bearing amphibolite at crustal levels. Radiogenic isotopes would probably not show the involvement of crustal rocks, because there is no evidence for the presence of crustal rocks old enough to have evolved radiogenic isotopic composition. Sr isotopic compositions of two amphibolites from northern Luzon, which have been carried up by a pluton, have ⁸⁷Sr/⁸⁶Sr of about 0.7039 to 0.7040 (U. Knittel, unpub. data, 1992).

It is also possible that no single process is responsible for the genesis of Pinatubo dacite but that several processes have worked in combination, such as extensive interaction between mantle-derived magma and crustal materials (for example, the MASH model of Hildreth and Moorbath, 1988). This model requires an efficient transfer of Be from the basalt to the dacite to account for the observed ¹⁰Be.

Tentatively we suggest the following petrogenetic model for Mount Pinatubo. Initially Pinatubo was an andesitic and dacitic stratovolcano (Newhall and others, this volume), built by lavas derived from primitive basalt generated by partial melting of the mantle wedge. Differentiation of the basalt is assumed to have taken place at the boundary between mantle and crust. Upon activation of the Iba fracture zone, deep faults provided conduits that allowed the primitive basalt to reach crustal levels. Due to their high density, these basalts did not reach the surface. During storage (and differentiation) of the basalt at crustal levels, the heat transfer to the host rocks caused them to melt partially. Possibly these anatectic melts mixed with differentiated melts derived from the basalt, to result in a homogeneous dacitic magma. Repeated injection of new primitive basalt into this reservoir may have triggered repeated violent eruptions.

The problem of the genesis of two texturally different types of dacite, phenocryst-rich and phenocryst-poor ones, cannot be solved in this model. Probably their genesis is related to the eruption dynamics.

MAGMA MIXING

The presence of olivine xenocrysts in dacite suggest that a small amount of mantle-derived basalt was mixed into the dacite; mixing of more substantial amounts of basalt with dacite resulted in the formation of andesite. (Pallister and others, 1992, calculated that a 40:60 mixture of basalt and dacite closely approximates the composition of the andesite).

Alternative sources for the olivine in dacite are unlikely: olivines from mantle peridotite and plutonic rocks typically have lower CaO content (<0.10 wt% CaO; Simkin and Smith, 1970) than those from Mount Pinatubo (0.19-0.25 wt% CaO; table 1), which are typical for volcanic rocks. Furthermore, olivines in plutonic mafic/ultramafic rocks (including metamorphic mantle peridotite) commonly lack tiny chromite inclusions, because this mineral forms discrete grains in deep-seated rocks, whereas chromite inclusions are common in olivine from primitive volcanic rocks. The ophiolitic basalts also are unlikely sources for the olivine xenocrysts because those basalts are mostly nonporphyritic, and clinopyroxene dominates over olivine among the phenocrysts (Evans and others, 1991). Hence, we would expect to see clinopyroxene xenocrysts also, if ophiolitic basalts were the source of xenocrystic olivine.

Therefore, the primitive basalts that were discovered as inclusions in andesite are the most likely source of the olivine xenocrysts (Knittel and others, 1991; Pallister and others, 1992). This conclusion is supported by the similar composition of the olivines and their chromite inclusions in dacite and basalt.

The presence of slightly contaminated basalt (with a few large quartz, amphibole, and plagioclase xenocrysts) in relatively homogeneous hybrid andesite, and the presence of olivine in dacite, suggest that mixing occurred in several stages. Following Koyaguchi (1986), we suggest that basalt intruding the magma reservoir ponded below dacite. Because of the faster convective overturn of the basalt (due to its lower viscosity), small batches of dacite were successively entrained into the basalt to result ultimately in the formation of the hybrid andesite. This andesite may have started vesiculation (high water content is indicated by the high amphibole content) and thereby reduced its density. This less-dense andesite may have become gravitationally unstable and started to rise. Thus, it disrupted the stratification of the magma chamber and trapped some basalt from the boundary layer between the hybrid andesite and the basalt. At the margins of the rising "diapir" some andesite was mixed into the dacite to produce the banded and mingled dacite. Some of this mixture was subsequently sufficiently diluted so that only the olivine xenocrysts, being isolated by amphibole rims, are still recognizable in dacite.

The evidence for the involvement of a basaltic component and mixing processes led Pallister and others (1992) to suggest that the eruption may have been triggered by injection of basalt into a reservoir containing dacitic melt. The evidence discussed above requires that some time must have elapsed between the injection of basalt into the reservoir and the eruption, in order for the andesite to have formed. It has been suggested that quartz may not survive a substantial period of time in disequilibrium with basalt. However, the clinopyroxene corona may have efficiently isolated the quartz from the basalt. Therefore, in view of the lack of other time constraints, we suggest that the events leading to the 1991 eruption were triggered by the July 16, 1990, earthquake. Influx of basalt into the dacite reservoir may initially have led only to increased solfataric activity in August 1990 (PHIVOLCS, 1990).

SEDIMENT INPUT

¹⁰Be contents of Pinatubo samples are high, 1.96x10⁶ and 2.77x10⁶ atoms of ¹⁰Be per gram of sample. Such high contents have been found outside arc settings only in rare cases where secondary, mostly atmospheric, alteration was likely (Tera and others, 1986). For the Pinatubo samples investigated here, such contamination can be ruled out because the samples were collected within a few days after the eruption. ¹⁰Be contamination via some kind of hydrothermal exchange with meteoric water may be possible, but is very unlikely in view of the huge volume of the Pinatubo eruption and the very high water to rock ratios required (Tera and others, 1986). Pristine ¹⁰Be content of volcanic rocks has been attributed to involvement of subducted sediments in the melting process (Tera and others, 1986; Morris and others, 1990).

Most of the western Pacific arc volcanoes are characterized by relatively low ¹⁰Be, as in Japan (mostly less than 0.7x10⁶ atoms ¹⁰Be/g), the Marianas (0.1-0.5x10⁶ atoms ¹⁰Be/g), and the Sunda arc (0.1-0.6x10⁶ atoms ¹⁰Be/g) (fig. 18) (Tera and others, 1986; Woodhead and Fraser, 1985). The high values at Pinatubo suggest either that sediment is particularly rapidly recycled into the source region of Mount Pinatubo or that the magma contains a relatively large sediment component. The slight shift towards elevated ⁸⁷Sr/⁸⁶Sr relative to ¹⁴³Nd/¹⁴⁴Nd (see McCulloch and others, 1980), and the radiogenic Pb, suggest involvement of sediments and igneous rock that have experienced seawater alteration.

Figure 18. ¹⁰Be content in lavas from several arcs of the western Pacific. Data sources: Pinatubo-this study; other data-Tera and others, 1986.

While the data leave little doubt about the presence of a recycled sedimentary component, the interpretation of these data is not straightforward, because the basalt found as inclusion in the dome andesite is not a potential precursor of the dacite. Hence the dacite may be derived from other sources, such as lower crustal sources, in which case the ¹⁰Be must have been transferred from the basalt to the dacite by fluid phases.

On a more regional scale, the presence of ¹⁰Be in the Pinatubo magmatic system supports the conclusion of Hayes and Lewis (1984) that sediment is subducted along the Manila trench. Furthermore, the presence of ¹⁰Be in the Pinatubo pumice may be considered as evidence that the Manila trench subduction system is still active, despite limited seismic activity.

ORIGIN OF THE SULFUR ENRICHMENT

The Pinatubo dacite, with a S content of 1,500 to 2,400 ppm, is extraordinarily enriched in S as compared with arc volcanic rocks, which typically contain less than 40 ppm S (Ueda and Sakai, 1984). The isotopic composition of this S is, however, not unusual for interoceanic arc magmas (fig. 19). All the Pinatubo ³⁴S values are characterized by much higher ³⁴S/³²S ratios than those reported for tholeiitic ocean ridge and ocean island basalt (³⁴S = -0.5 to +1.0) (Harmon and Hoefs, 1986; Sakai and others, 1982, 1984). They are, however, typical of other interoceanic arc volcanics, which are characterized by a large range of ³⁴S (Ueda and Sakai, 1984; Woodhead and others, 1987) (fig. 19).

Figure 19. ³⁴S values of dacites and andesites from Pinatubo compared to other intraoceanic arc magmas (modified from Woodhead and others, 1987).

The two potential sources usually proposed for this S enrichment are S-rich crustal sediments such as marine evaporites or hydrothermal sulfide deposits in subducted oceanic crust. If the sulfur enrichment observed in the Pinatubo dacite is a consequence of the assimilation of evaporite beds by the magma on its way to the surface, the ³⁴S should reflect much higher values, close to those reported for Oligocene evaporites of around 20 to 22 (Claypool and others, 1980). Moreover, there is no geologic evidence for the presence of evaporitic beds in the basement of the volcano or on the Luzon arc. It is therefore unlikely that contamination by evaporites is the source of this unusual S enrichment. The relatively low ⁸⁷Sr/⁸⁶Sr isotopic ratios also preclude significant contributions from marine evaporites. On the other hand, sulfur of magmatic origin occurs in anhydrite in hydrothermally altered volcanic rocks in the basement of Mount Pinatubo, so crustal contamination as a mechanism of S enrichment cannot be excluded (Hattori, this volume).

Alternatively, hydrothermal sulfide deposits, generated on the oceanic crust, could have been subducted along the Manila trench and could be responsible for the unusual S enrichment. The isotopic compositions of these hydrothermal sulfide range from ³⁴S = +4.3 to +10.7, being generally between +6 to +8 (Hallbach and others, 1989; Kusakabe and others, 1990). These relatively heavy S compositions are compatible with the ³⁴S observed for the Pinatubo dacites.

Mass balance calculations suggest that during the eruption about half of the total S content of the magma was released as a gas phase and the remainder is essentially present as anhydrite phenocrysts (Westrich and Gerlach, 1992). In consequence, determination of the sulfur isotopic composition of the gas phase is essential to evaluate the ³⁴S of the bulk magma. If the sulfate adsorbed on two ash samples (Pin12 and Pin14) is representative of the gas phase, the isotopic composition of leachates suggests that the vapor had roughly the same composition as the S left in the magma as anhydrite. Thus, on the basis of these S isotopic compositions, anhydrite breakdown as the source of stratospheric S cannot be confirmed or ruled out.

CONCLUSIONS

The mineralogical, geochemical, and isotopic data presented reveal no unusual features that may be connected with the extreme sulfur enrichment of the 1991 Pinatubo eruption products. While there is evidence for the involvement of recycled crustal material in the generation of the Pinatubo dacites, the extent of this contribution seems to be of similar magnitude as in other volcanoes of the Bataan segment. In particular, there is no evidence for the involvement of marine evaporites. Hence, the process leading to this unusual sulfur enrichment cannot be detected by classical petrological studies. In consequence, it would be extremely difficult to detect a similar sulfur-rich magma from old eruptions if anhydrite is completely removed, because no uncommon features are displayed except for the presence of anhydrite and a relatively high f_O2. Clearly this leaves open the possibility that sulfur-rich eruptions are far more frequent than hitherto expected.

A few volcanoes are found to emit excess amounts of sulfur into the atmosphere. Basaltic magma as a source for this excess sulfur has repeatedly been invoked (see Andres and others, 1991). In the case of the 1991 Pinatubo eruption, there is clear evidence for the involvement of a basaltic component, the volume of which, however, is unknown. Similar basalt may have been added to the magma reservoir during previous mixing episodes, well before the June 12-15 eruption, but its role as a source for sulfur is doubtful for several reasons. First, the high f_O2 of the dacite is not compatible with the exsolution of volatiles from a reduced basaltic melt, as also suggested by Westrich and Gerlach (1992). Second, the homogeneous distribution of anhydrite throughout a minimum of 5 km³ of dacite is difficult to understand if the S had been recently introduced as a volatile phase from the bottom of the dacite magma. A homogeneous distribution of the anhydrite is more plausible if it formed following saturation of the melt in sulfate.

ACKNOWLEDGMENTS

We thank Ray Punongbayan and PHIVOLCS scientists for support in the field. Data collection in Aachen and participation in the International Scientific Conference on Mount Pinatubo in Manila was made possible by grants of Deutsche Forschungsgemeinschaft (grant Kn237/5-1 and travel grant). Fieldwork by Oles was largely sponsored by Professor H. Förster. Fieldwork by Bernard was supported by grants from Fonds National de la Recherche Scientifique (Belgium). M.D. Feigenson kindly determined the Sr and Nd isotopic composition of the two samples analyzed for ¹⁰Be. We thank J.P. Mennessier (Brussels) for his help processing the anhydrite and lavas for chemical analysis, J.C. Duchesne and G. Bologne (Liège) for X-ray fluorescence analyses, and S. Wauthier (Louvain La Neuve) for the electron microprobe analyses. We also acknowledge preprints of other contributions to the Pinatubo volume that we received from J.S. Pallister, P.R. Castillo, M.A. McKibben, and A. Imai. C. Arcilla commented on several early versions of the manuscript. Finally we wish to thank our reviewers, T.M. Gerlach and M.J. Defant, largely for encouragement, and J.S. Pallister for a very detailed and constructive review; largely due to his review, the number of errors, misinterpretations, and omissions was kept at a minimum.

REFERENCES CITED

Andres, R.J., Rose, W.I., Kyle, P.R., de Silva, S., Francis, P., Gardeweg, M., and Moreno Roa, H., 1991, Excessive sulfur dioxide emissions from Chilean volcanoes: Journal of Volcanology and Geothermal Research, v. 46, p. 323-329.

Arculus, R.J., Johnson, R.W., Chappell, B.W., McKee, C.O., and Sakai, H., 1983, Ophiolite-contaminated andesites, trachybasalts, and cognate inclusions of Mt. Lamington, Papua New Guinea anhydrite-amphibole-bearing lavas and the 1951 cumulodome: Journal of Volcanology and Geothermal Research, v. 18, p. 21-247.

Bau, M., and Knittel, U., 1993, Significance of slab-derived partial melts and aqueous fluids for the genesis of tholeiitic and calc-alkaline island-arc basalts: Evidence from Mt. Arayat, Philippines: Chemical Geology, v. 105, p. 233-251.

Bernard, A., Demaiffe, D., Mattielli, N., and Punongbayan, R.S., 1991, Anhydrite-bearing pumices from Mount Pinatubo: Further evidence for the existence of sulphur-rich silicic magmas: Nature, v. 354, p. 139-140.

Bluth, G.J.S., Doiron, S.D., Schnetzler, C.C., Krueger, A.J., and Walter, L.S., 1992, Global tracking of the SO₂ clouds from the June, 1991 Mount Pinatubo eruptions: Geophysical Research Letters, v. 19, p. 151-154.

Brown, L., 1984, Applications of accelerator mass spectrometry: Annual Review of Earth and Planetary Science, v. 12, p. 39-69.

Cardwell, R.K., Isacks, B.L., and Karig, D.E., 1980, The spatial distribution of earthquakes, focal mechanism solutions, and subducted lithosphere in the Philippine and Northeastern Indonesian Islands, in Hayes, D.E., ed., The tectonic and geologic evolution of Southeast Asia: American Geophysical Union Monograph 23, p. 1-35.

Castillo, P.R., and Punongbayan, R.S., this volume, Petrology and Sr, Nd, and Pb isotopic geochemistry of Mount Pinatubo volcanic rocks.

Catanzaro, E.J., Murphy, T.J., Shields, W.R., and Garner, E.L., 1968, Absolute isotopic abundance ratios of common, equal-atom, and radiogenic lead isotope standards: Journal of Research of the National Bureau of Standards, v. 72A, p. 261-267.

Chung, S.-l., and Sun, S.-s., 1992, A new genetic model for the East Taiwan Ophiolite and its implications for Dupal domains in the Northern Hemisphere: Tectonophysics, v. 109, p. 133-145.

Claypool, G.E., Holser, W.T., Kaplan, I.R., Sakai, H., and Zak, T., 1980, The age curves of sulfur and oxygen isotopes in marine sulfate and their mutual interpretation: Chemical Geology, v. 28, p. 199-260.

de Boer, J., Odom, L.A., Ragland, P.C., Snider, F.G., and Tilford, N.R., 1980, The Bataan Orogene: Eastward subduction, tectonic rotations, and volcanism in the western Pacifics (Philippines): Tectonophysics, v. 67, p. 251-282.

Defant, M.J., 1985, The potential origin of the potassium depth relationship in the Bataan orogene, the Philippines: Tallahassee, Florida State University, Ph.D. dissertation, 622 p.

Defant, M.J., and Drummond, M.S., 1990, Derivation of some modern arc magmas by melting of young subducted lithosphere: Nature, v. 347, p. 662-664.

------1993, Mount St. Helens: Potential example of the partial melting of the subducted lithosphere in a volcanic arc: Geology, v. 21, p. 547-550.

Defant, M.J., Jacques, D., Maury, R.C., de Boer, J.Z., and Joron, J.L., 1989, Geochemistry of the Luzon arc, Philippines: Geological Society of America Bulletin, v. 101, p. 663-672.

Defant, M.J., Maury, R.C., Joron, J.-L., Feigenson, M.D., Leterrier, J., Bellon, H., Jacques, D., and Richard, M., 1990, The geochemistry and tectonic setting of the northern section of the Luzon arc (the Philippines and Taiwan): Tectonophysics, v. 183, p. 187-205.

Defant, M.J., Maury, R.C., Ripley, E.M., Feigenson, M.D., and Jacques, D., 1991, An example of island-arc petrogenesis: Geochemistry and petrology of the southern Luzon Arc, Philippines: Journal of Petrology, v. 32, p. 455-500.

Dick, H.J.B., and Bullen, T., 1984, Chromian spinel as a petrogenetic indicator in abyssal and alpine-type peridotites and spatially associated lavas: Contributions to Mineralogy and Petrology, v. 86, p. 54-76.

Drummond, M.S., and Defant, M.J., 1990, A model for trondhjemite- tonalite-dacite genesis and crustal growth via slab-melting: Archean to modern comparisons: Journal of Geophysical Research, v. 95, p. 21503-21521.

Evans, C.A., Casteneda, G., and Franco, H., 1991, Geochemical complexities preserved in the volcanic rocks of the Zambales Ophiolite, Philippines: Journal of Geophysical Research, v. 96, p. 16251-16262.

Falloon, T.J., Green, D.H., Hatton, C.J., and Harris, K.L., 1988, Anhydrous partial melting of a fertile and depleted peridotite from 2 to 30 kb and application to basalt petrogenesis: Journal of Petrology, v. 29, p. 1257-1282.

Förster, H., Oles, D., Knittel, U., Defant, M.C., and Torres, R.C., 1990, The Macolod Corridor: A rift crossing the Philippine island arc: Tectonophysics, v. 183, p. 265-271.

Fournelle, J.H., 1991, Anhydrite and sulfide in pumices from the 15 June 1991 eruption of Mt. Pinatubo; initial examination [abs.]: Eos, Transactions, American Geophysical Union, v. 72, no. 44, p. 68.

Geschwind, C.-H., and Rutherford, M.J., 1992, Cummingtonite and the evolution of the Mount St. Helens (Washington) magma system: An experimental study: Geology, v. 20, p. 1011-1014.

Giese, U., Knittel, U., and Kramm, U., 1986, The Paracale Intrusion: Geologic setting and petrogenesis of a trondhjemite intrusion in the Philippine island arc: Journal of Southeast Asian Earth Sciences, v. 1, p. 235-245.

Gill, J.B., 1981, Orogenic andesites and plate tectonics: Heidelberg, Springer, 390 p.

Graham, I.J., and Worthington, 1988, Petrogenesis of Tauhara dacite (Taupo Volcanic Zone, New Zealand)--evidence for magma mixing between high-alumina andesite and rhyolite: Journal of Volcanology and Geothermal Research, v. 35, p. 279-294.

Hallbach, P., Nakamura, K.-i., Wahsner, M., Lange, J., Sakai,H., Kaselitz, L., Hansen, R.-D., Yamano, M., Post, J., Prause, B., Seifert, R., Michaelis, W., Teichmann, F., Kinoshita, M., Marten, A., Ishibashi, J., Czerwinski, S., and Blum, N., 1989, Probable modern analogue of Kuroko-type massive sulfide deposits in the Okinawa Trough back-arc basin: Nature, v. 338, p. 496-499.

Harmon, R.S., and Hoefs, J., 1986, S-isotope relationships in late Cenozoic destructive plate margin and continental intra-plate volcanic rocks: Terra Cognita, v. 6, p. 182.

Hattori, K., this volume, Occurrence and origin of sulfide and sulfate in the 1991 Mount Pinatubo eruption products.

Hayes, D.E., and Lewis, S.D., 1984, A geophysical study of the Manila Trench, Luzon, Philippines. 1, Crustal structure, gravity, and regional tectonic evolution: Journal of Geophysical Research, v. 89, p. 9171-9195.

Helz, R.T., 1976, Phase relations of basalt in their melting ranges at P _H2O = 5 Kb. Part II, Melt compositions: Journal of Petrology, v. 17, p. 139-193.

Hildreth, W., and Moorbath, S., 1988, Crustal contributions to arc magmatism in the Andes of central Chile: Contributions to Mineralogy and Petrology, v. 98, p. 455-489.

Hofman, D.J., Oltmans, S.J., Harris, J.M., Solomon, S., Deshler, T., and Johnson, B.J., 1992, Observation and possible causes of new ozone depletion in Antartica in 1991: Nature, v. 359, p. 283-287.

Imai, A., Listanco, E.L, and Fujii, T., this volume, Highly oxidized and sulfur-rich dacitic magma of Mount Pinatubo: Implication for metallogenesis of porphyry copper mineralization in the Western Luzon arc.

Irving, A.J., and Frey, F.A., 1978, Distribution of trace elements between garnet megacrysts and host volcanic liquids of kimberlitic to rhyolitic composition: Geochimica and Cosmochimica Acta, v. 42, p. 771-787.

Jahn, B.-m., 1986, Mid-ocean ridge or marginal basin origin of the East Taiwan Ophiolite: Chemical and isotopic evidence: Contributions to Mineralogy and Petrology, v. 92, p. 194-206.

Kay, R.W., 1978, Aleutian magnesian andesites: Melts from subducted Pacific ocean crust: Journal of Volcanology and Geothermal Research, v. 4, p. 117-132.

Kerr, R.A., 1993, Pinatubo global cooling on target: Science, v. 259, p. 594.

Klein, J., and Middleton, R., 1984, Accelerator mass spectrometry at the University of Pennsylvania: Nuclear Instruments and Methods, v. B5, p. 129-133.

Knittel., U., and Defant, M.J., 1988, Sr isotopic and trace element variations in Oligocene to Recent igneous rocks from the Philippine island arc: Evidence for recent enrichment in the sub-Philippine mantle: Earth and Planetary Science Letters, v. 87, p. 87-99.

Knittel, U., Defant, M.J., and Razcek, I., 1988, Recent enrichment in the source region of arc magmas from Luzon island, Philippines: Sr and Nd isotopic evidence: Geology, v. 16, p. 73-76.

Knittel, U., Hegner, E., Satir, M., Oles, D., and Förster, H., 1992, ⁸⁷Sr/⁸⁶Sr and¹⁴³Nd/¹⁴⁴Nd isotopic ratios in rift basalts from the Taiwan-Luzon arc: implications for the composition of the sub-arc mantle: European Journal of Mineralogy, Supplement, v. 1, p. 150.

Knittel, U., and Oles, D., 1995, Basaltic volcanism associated with extensional tectonics in the Taiwan-Luzon island arc: Evidence for non-depleted sources and subduction zone enrichment, in Smellie, J.L., ed., Volcanism Associated with extension at Consuming Plate Margins: Geological Society Special Publication 81, p. 77-93.

Knittel, U., Oles, D., Förster, H., and Punongbayan, R.S., 1991, Pinatubo pumice contains anhydrite: Eos, Transactions of the American Geophysical Union, v. 72, p. 522.

Knittel, U., Trudu, A., Winter, W., Yang, T.F., and Gray, C., in press, Volcanism above a subducted extinct spreading center: A reconnaissance study of the North Luzon Segment of the Taiwan-Luzon Volcanic Arc (Philippines): Journal of Southeast Asian Earth Sciences.

Knittel-Weber, C., and Knittel, U., 1990, Petrology and genesis of the volcanic rocks on the eastern flank of Mount Malinao, Bicol arc (southern Luzon, Philippines): Journal of Southeast Asian Earth Sciences, v. 4, p. 267-280.

Koyaguchi, T., 1986, Evidence for two-stage mixing in magmatic inclusions and rhyolite lava domes on Niijima Island, Japan: Journal of Volcanology and Geothermal Research, v. 29, p. 71-98.

Kusakabe, M., Mayeda, S., and Nakamura, E., 1990, S, O, and Sr isotope systematics of active vent materials from the Mariana backarc basin spreading axis at 18°N: Earth and Planetary Science Letters, v. 100, p. 275-282.

Leeman, W.P., Smith, D.R., Hildreth, W., Palacz, Z., and Rogers, N., 1990, Compositional diversity of Late Cenozoic basalts in a transect across the southern Washington Cascades: Implications for subduction zone magmatism: Journal of Geophysical Research, v. 95, p. 19561-19582.

Luhr, J.F., 1990, Experimental phase relations of water- and sulfur-saturated arc magmas and the 1982 eruptions of El Chichón Volcano: Journal of Petrology, v. 31, p. 1071-1114.

Luhr, J.F., Carmichael, I.S.E., and Varekamp, J.C., 1984, The 1982 eruptions of El Chichón Volcano, Chiapas, Mexico: Mineralogy and petrology of the anhydrite-bearing pumices: Journal of Volcanology and Geothermal Research, v. 23, p. 69-108.

Manhès, G., Minster, J.F., and Allègre, C.J., 1978, Comparative uranium-thorium-lead and rubidium-strontium study of the St. Severin amphoterite: Consequences for early solar system chronology: Earth and Planetary Science Letters, v. 39, p. 14-24.

Martin, H., 1987, Petrogenesis of Archaean trondhjemites, tonalites, and granodiorites from Eastern Finland: Major and trace element geochemistry: Journal of Petrology, v. 28, p. 921-953.

McCormick, M.P., and Veiga, R.E., 1992, SAGE II measurements of early Pinatubo aerosols: Geophysical Research Letters, v. 19, p. 155-158.

McCulloch, M.T., Gregory, R.T., Wasserburg, G.J., and Taylor, H.P., 1980, A neodymium, strontium, and oxygen isotopic study of the Cretaceous Samail ophiolite and implications for the petrogenesis and seawater-hydrothermal alteration of oceanic crust: Earth and Planetary Science Letters, v. 46, p. 201-211.

McDermott, F., Defant, J.J., Hawkesworth, C.J., Maury, R.C., and Joron, J.L., 1993, Isotope and trace element evidence for three component mixing in the genesis of the North Luzon arc lavas (Philippines): Contributions to Mineralogy and Petrology, v. 113, p. 9-23.

McKibben, M.A., Eldridge, C.S., and Reyes, A.G., this volume, Sulfur isotopic systematics of the June 1991 Mount Pinatubo eruptions: A SHRIMP ion microprobe study.

Miklius, A., Flower, M.F.J., Huijsmans, J.P.P., Mukasa, S.B., and Castillo, P., 1991, Geochemistry of lavas from Taal volcano, southwestern Luzon, Philippines: Evidence for multiple magma supply systems and mantle source heterogeneity: Journal of Petrology, v. 32, p. 593-627.

Morris, J.D., Leeman, W.P., and Tera, F., 1990, The subducted component in island arc lavas: Constraints from Be isotopes and B-Be systematics: Nature, v. 344, p. 31-36.

Mukasa, S.B., Flower, M.F.J., Miklius, A., 1994, The Nd-, Sr-, and Pb-isotopic character of lavas from Taal, Laguna de Bay and Arayat Volcanoes, Southwestern Luzon, Philippines: implications for arc petrogenesis: Tectonophysics, v. 235, p. 205-221.

Mukasa, S.B., McCabe, R., and Gill, J.B., 1987, Pb-isotopic compositions of volcanic rocks in the West and East Philippine island arc: Presence of the Dupal isotopic anomaly: Earth and Planetary Science Letters, v. 84, p. 153-164.

Nakamura, H., 1974, Determination of REE, Ba, Mg, Na, and K in carbonaceous and ordinary chondrites: Geochimica et Cosmochimica Acta, v. 38, p. 757-775.

Newhall, C.G., Daag, A.S., Delfin, F.G., Jr., Hoblitt, R.P., McGeehin, J., Pallister, J.S., Regalado, M.T.M., Rubin, M., Tamayo, R.A., Jr., Tubianosa, B., and Umbal, J.V., this volume, Eruptive history of Mount Pinatubo.

Nicholls, I.A., and Whitford, D.J., 1976, Primary magmas associated with Quaternary volcanism in the western Sunda arc, Indonesia, in Johnson, R.W., ed., Volcanism in Australasia: Amsterdam, Elsevier, p. 77-90.

Nye, C.J., and Reid, M.R., 1986, Geochemistry of primary and least fractionated lavas from Okmok Volcano, central Aleutians: Implications for arc magma genesis: Journal of Geophysical Research, v. 91, p. 10271-10287.

Oles, D., 1988, Das San Pablo-Vulkangebiet (Philippinen): Graben-vulkanismus im Luzon-Inselbogen (The San Pablo Volcanic Field: volcanism in a horst and graben setting in the Luzon volcanic arc): Aachen, Rheinisch Westfälische Technische Hochschule, Ph.D. dissertation, 156 p.

Pallister, J.S., Hoblitt, R.P., Meeker, G.P., Knight, R.J., and Siems, D.F., this volume, Magma mixing at Mount Pinatubo: Petrographic and chemical evidence from the 1991 deposits.

Pallister, J.S., Hoblitt, R.P., and Reyes, A.G., 1992, A basalt trigger for the 1991 eruptions of Pinatubo volcano? Nature, v. 356, p. 426-428.

PHIVOLCS, 1990, Mount Pinatubo not erupting, PHIVOLCS Observer, July-September 1990, v. 6, no. 3, p. 6.

Pinatubo Volcano Observatory Team, 1991, Lessons from a major eruption: Mount Pinatubo, Philippines: Eos, Transactions of the American Geophysical Union, v. 72, p. 545-555.

Rapp, R.P., Watson, E.B., and Miller, C.F., 1991, Partial melting of amphibolite/eclogite and the origin of Archean trondhjemites and tonalites: Precambrian Research, v. 51, p. 1-25.

Reid, F.W., and Cole, J.W., 1983, Origin of dacites of Taupo Volcanic Zone, New Zealand: Journal of Volcanology and Geothermal Research, v. 18, p. 191-214.

Richard, P., Shimizu, N., and Allègre, C.J., 1976, ¹⁴³Nd/¹⁴⁶Nd, a natural tracer: an application to oceanic basalts: Earth and Planetary Science Letters, v. 31, p. 269-278.

Roeder, P.L., and Emslie, R.F., 1970, Olivine-liquid equilibrium: Contributions to Mineralogy and Petrology, v. 29, p. 275-289.

Rutherford, M.J., 1993, Experimental petrology applied to volcanic processes: Eos, Transactions of the American Geophysical Union, v. 74, p. 49-52.

Rutherford, M.J., and Devine, J.D., 1991, Pre-eruption conditions and volatiles in the 1991 Pinatubo magma [abs.]: Eos, Transactions, American Geophysical Union, v. 72, no. 44, p. 62.

------this volume, Preeruption pressure-temperature conditions and volatiles in the 1991 dacitic magma of Mount Pinatubo.

Sabit, J.P., 1992, Pinatubo volcano's 1991 eruptions: Unpublished manuscript distributed at the International Scientific Conference on Mt. Pinatubo, Manila, Department of Foreign Affairs, May 1992.

Sabit, J.P., Pigtain, R.C., and de la Cruz, E.G., this volume, The west-side story: Observations of the 1991 Mount Pinatubo eruptions from the west.

Sakai, H., Casadevall, T.J., and Moore, J.G., 1982, Chemistry and isotope ratios of sulphur in basalts and volcanic gases at Kilauea volcano, Hawaii: Geochimica et Cosmochimica Acta, v. 46, p. 729-738.

Sakai, H., Des Marais, D.J., Ueda, A., and Moore, J.G., 1984, Concentration and isotopic ratios of carbon, nitrogen, and sulphur in ocean-floor basalts: Geochimica et Cosmochimica Acta, v. 48, p. 2433-2441.

Scott, W.E., Hoblitt, R.P., Torres, R.C., Self, S, Martinez, M.L., and Nillos, T., Jr., this volume, Pyroclastic flows of the June 15, 1991, climactic eruption of Mount Pinatubo.

Sigurdsson, H., Devine, J.D., and Davis, A.N., 1985, The petrologic estimation of volcanic degassing: Jokull, v. 35, p. 1-8.

Simkin, T., and Smith, J.V., 1970, Minor element distribution in olivine: Journal of Geology, v. 78, p. 304-325.

Smith, D.R., and Leeman, W.P., 1987, Petrogenesis of Mount St. Helens dacitic magmas: Journal of Geophysical Research, v. 92, p. 10313-10334.

Stowe, L.L., Carey, R.M., and Pellegrino, P.P., 1992, Monitoring the Mt. Pinatubo layer with NOAA/11 AVHRR data: Geophysical Research Letters, v. 19, p. 159-162.

Tera, F., Brown, L., Morris, J., Sacks, I.S., Klein, J., and Middleton, R., 1986, Sediment incorporation in island-arc magmas: Inferences from ¹⁰Be: Geochimica et Cosmochimica Acta, v. 50, p. 535-550.

Ueda, A., and Sakai, H., 1984, Sulfur isotope study of Quaternary volcanic rocks from the Japanese Island Arc: Geochimica et Cosmochimica Acta, v. 48, p. 1837-1848.

Vidal, Ph., Dupuy, C., Maury, R., and Richard, M., 1989, Mantle metasomatism above subduction zones: Trace-element and radiogenic isotope characteristics of peridotite xenoliths from Bataan Island (Philippines): Geology, v. 17, p. 1115-1118.

Voss, F., 1971, Die geologische und morphologische Struktur des aktivsten Vulkangebietes der Philippinen (The geologic and morphologic structure of the most active volcanic field in the Philippines): Die Erde, v. 102, p. 308-316.

Wasserburg, G.J., Jacobsen, S.B., DePaolo, D.J., McCulloch, M.T., and Wen, T., 1981, Precise determination of Sm/Nd ratios, Sm and Nd isotopic abundances in standard solutions: Geochimica et Cosmochimica Acta, v. 45, p. 2311-2323.

Weis, D., Demaiffe, D., Cauët, S., and Javoy, M., 1987, Sr, Nd, O and H isotopic ratios in Ascension lavas and plutonic inclusions: Cogenetic origin: Earth and Planetary Science Letters, v. 82, p. 255-268.

Westrich, H.R., and Gerlach, T.M., 1992, Magmatic gas source for the stratospheric SO 2 cloud from the June 15, 1991, eruption of Mount Pinatubo: Geology, v. 20, p. 867-870.

Wolfe, J.A., and Self, S., 1983, Structural lineaments and Neogene volcanism in southwestern Luzon, in Hayes, D.E., ed., The tectonic and geologic evolution of Southeast Asian seas and islands, Part 2: American Geophysical Union, Geophysical Monograph 27, p. 157-172.

Woodhead, J.D., and Fraser, D.G., 1985, Pb, Sr and ¹⁰Be studies of volcanic rocks from the northern Mariana Islands--Implications for magma genesis and crustal recycling in the Western Pacific: Geochimica et Cosmochimica Acta., v. 49, p. 1925-1930.

Woodhead, J.D., Harmon, R.S., and Fraser, D.G., 1987, O, S, Sr and Pb isotope variations in volcanic rocks from the northern Mariana Islands: Implications for crustal recycling in intra-oceanic arcs: Earth and Planetary Science Letters, v. 83, p. 39-52.

Yan, X-H., Ho, C-R, Zheng, Q., and Klemas, V., 1992, Temperature and size variabilities of the Western Pacific warm pool: Science, v. 258, p. 1643-1645.

FIRE and MUD Contents

PHIVOLCS | University of Washington Press | U.S.Geological Survey

This page is <http://pubs.usgs.gov/pinatubo/bernard/>
Contact: Chris Newhall
Last updated 06.10.99