See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/225125197

Geochemistry of silicic magmas in the Macolod Corridor, SW Luzon,

Philippines: Evidence of distinct, mantle-derived, crustal sources for silicic
magmas

Article  in  Contributions to Mineralogy and Petrology · March 2006

DOI: 10.1007/s00410-005-0050-7

CITATIONS READS

31 3,664

8 authors, including:

Thomas Vogel Tim Flood

Michigan State University St. Norbert College
98 PUBLICATIONS   1,850 CITATIONS    17 PUBLICATIONS   90 CITATIONS   

SEE PROFILE SEE PROFILE

Raymond Patrick Maximo Maria Carmencita B. Arpa

Université Libre de Bruxelles 17 PUBLICATIONS   116 CITATIONS   
7 PUBLICATIONS   38 CITATIONS   
SEE PROFILE
SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Evolution of silicic magmas in Guanacaste Province, Costa Rica. View project

Geochemical studies of thermal springs on active volcanoes in the Philippines. View project

All content following this page was uploaded by Thomas Vogel on 17 July 2015.

The user has requested enhancement of the downloaded file.

Contrib Mineral Petrol (2006) 151: 267–281
DOI 10.1007/s00410-005-0050-7

O R I GI N A L P A P E R

Thomas A. Vogel Æ Timothy P. Flood

Lina C. Patino Æ Melissa S. Wilmot
Raymond Patrick R. Maximo Æ Carmencita B. Arpa
Carlo A. Arcilla Æ James A. Stimac

Geochemistry of silicic magmas in the Macolod Corridor, SW Luzon,

Philippines: evidence of distinct, mantle-derived, crustal sources
for silicic magmas
Received: 28 June 2005 / Accepted: 15 November 2005 / Published online: 14 February 2006

Springer-Verlag 2006

Abstract Silicic volcanic deposits (>65 wt% SiO2), which of primitive basalts and andesites demonstrate that it is
occur as domes, lavas and pyroclastic deposits, are rela- difficult to produce high K2O/Na2O silicic magmas by
tively abundant in the Macolod Corridor, SW Luzon, fractional crystallization or partial melting of a low K2O/
Philippines. At Makiling stratovolcano, silicic domes Na2O source. However, recent melting experiments (Sis-
occur along the margins of the volcano and are chemically son et al., Contrib Mineral Petrol 148:635–661, 2005)
similar to the silicic lavas that comprise part of the vol- demonstrate that extreme fractional crystallization or
cano. Pyroclastic flows are associated with the Laguna de partial melting of K-rich basalts can produce these silicic
Bay Caldera and these are chemically distinct from the magmas. Our model for the generation of the silicic
domes and lavas at Makiling stratovolcano. As a whole, magmas in the Macolod Corridor requires partial melting
samples from the Laguna de Bay Caldera contain lower of mantle-derived, evolved, moderate to K-rich, crystal-
concentrations of MgO and higher concentrations of lized calc-alkaline magmas that ponded and crystallized in
Fe2O3(t) than the samples from domes and lavas. The the mid-crust. Major and trace element variations, along
Laguna de Bay samples are more enriched in incompatible with oxygen isotopes and ages of the deposits, are con-
trace elements. The silicic rocks from the domes, Makiling sistent with this model.
Volcano and Laguna de Bay Caldera all contain high
alkalis and high K2O/Na2O ratios. Melting experiments

Electronic Supplementary Material Supplementary material is

available for this article at http://dx.doi.org/10.1007/s00410-005- Introduction
0050-7
The production of silicic magmas in oceanic arc settings,
Communicated by T. L. Grove
without the involvement of evolved continental crust, has
T. A. Vogel (&) Æ L. C. Patino Æ M. S. Wilmot Æ C. B. Arpa been controversial. Models for the origin of silicic mag-
Department of Geological Sciences, Michigan State University, mas in areas with continental crust involve partial melting
East Lansing, MI 48824-1115, USA or assimilation of crustal rocks (DePaolo 1981; Cobbing
E-mail: vogel@msu.edu
Tel.: +1-517-3539029
and Pitcher 1983; White and Chappell 1983; Hildreth and
Fax: +1-517-3538787 Moorbath 1988; Vielzeuf and Holloway 1988). On the
contrary, silicic magmatism has been considered to be
T. P. Flood minor in oceanic arcs where evolved continental crust is
Department of Geology, St. Norbert College, 100 Grant Street, absent. However, recent works (Tamura and Tatsumi
De Pere, WI 54115-2099, USA
2002; Leat et al. 2003; Smith et al. 2003; Vogel et al. 2004)
R. P. R. Maximo Æ C. B. Arpa have shown that silicic volcanism can be significant in
Philippine Institute of Volcanology and Seismology, Quezon City, oceanic arcs without old, evolved, continental-like crust.
Philippines Workers propose that the generation of silicic magmas is
C. A. Arcilla due to the partial melting of, or melt extraction from,
University of the Philippines—Diliman, National Institute recently emplaced, mantle-derived, stalled (crystallized or
of Geological Sciences, Quezon City, Philippines partially crystallized), calc-alkaline magmas. Signifi-
cantly, the emplacement of silicic magmas in areas with-
J. A. Stimac
Unocal Philippines, Inc, 12th Floor, Citibank Tower,
out old, evolved crust represents the formation of new
1226 Makati City, Philippines continental crust.
268

The Philippine Archipelago is built upon oceanic arcs

with no old, evolved, continental crust, except for the Geologic setting
Palawan-Mindoro microcontentinal block to the
southwest (Knittel et al. 1988) (Fig. 1). Silicic volcanic Most of the Philippine Archipelago lies between two
deposits (>65 wt% SiO2) are relatively abundant in the opposing subduction systems (Fig. 1). At the latitude of
Macolod Corridor, SW Luzon, Philippines (Figs. 1, 2). Luzon Island, the Eurasian (Sunda) Plate is being sub-
They occur as domes on the western and southern flanks ducted eastward along the Manila Trench (Cardwell
of Makiling Volcano, as lava flows associated with et al. 1980), whereas the Philippine Sea Plate is under-
Makiling Volcano and as pyroclastic flows and related going a westward subduction along the Philippine
deposits erupted from Laguna de Bay Caldera (Fig. 3). Trench. The Luzon Arc is the product of eastward sub-
This paper focuses on the origin of the magmas that duction and has been divided into two segments, the
produced silicic domes (Fig. 3) and compares them to Bataan Arc to the north and the Mindoro Arc to the
the lavas at Makiling stratovolcano (J.B. Cruz, unpub- south based on distinctive geochemical and isotopic
lished thesis) as well as to the relatively large volume trends in recent volcanic rocks (Defant et al. 1988). The
pyroclastic deposits erupted from Laguna de Bay Macolod Corridor is a NE-trending, 80 km wide zone of
Caldera (Catane and Arpa 1998). These comparisons extensional faulting between these two segments of the
will be used to constrain genetic models. Luzon Arc (Sudo et al. 2001) that is nearly perpendicular

Fig. 1 Tectonic map of the

Philippine Archipelago,
showing the setting of the
Macolod Corridor (Bacolcol
2003). Close triangles are
volcanoes; hachures on faults
indicate downthrown block;
large arrows indicate opening
direction of Macolod Corridor
 269

Corridor (Pubellier et al. 2000; Aquino 2004) although

the general consensus is that it is a pull-apart rift zone.
Adjacent to the Macolod Corridor is a sequence of
Paleocene–Oligocene San Juan metavolcanics and me-
tasediments and early to mid-Miocene San Juan quartz
diorite (Oles 1991; Sudo et al. 2001).
Volcanic centers within the Macolod Corridor consist
of two large volcano–tectonic depressions (Taal and
Laguna de Bay Calderas), three stratovolcanoes
(Makiling, Cristo and Banahaw) and more than 200
small monogenetic volcanic scoria cones, domes and
maars (Förster et al. 1990) (Fig. 2). The Laguna de Bay
Caldera is a 200 km2 depression that forms the central
lobe of a large, three-fingered lake called Laguna de Bay
(Fig. 2). This caldera has erupted large volumes of
pyroclastic material, with reported ages from about 27
to about 50 ka (Catane and Arpa 1998; Arpa et al.
1999a; Catane et al. 2004). Although the volume of
erupted products has not been determined, the caldera is
15·30 km2 (Arpa et al. 1999b) and pyroclastic units
from this caldera have been observed in metro Manila,
40 km to the northwest. Abundant fresh outcrops are
observed in quarries in the immediate vicinity of the
caldera, but outcrops are poor, due to weathering, where
quarries are absent. Based on our chemical analyses of
pumice fragments, the pyroclastic deposits range in
composition from 53 to 69 wt% SiO2. The Makiling
Fig. 2 Simplified geologic map of the Macolod Corridor. Triangles volcanic complex is composed of lava flows, coalesced
are volcanoes or volcanic centers. Faults are represented by domes and pyroclastic deposits (Fig. 3), primarily of
hachured lines (hachures indicate the downthrown area), open basaltic andesite and andesite composition with lesser
arrows point to the inferred direction of minimum horizontal stress amounts of trachyandesite and dacite (J.B. Cruz,
and ongoing extension of the Macolod Corridor. Laguna de Bay
Caldera is the middle lobe of Laguna de Bay (Le Rouzic 1999) unpublished thesis).
The initiation of the most recent episodes of volca-
nism in the Macolod Corridor is not well constrained,
to the arc front (Figs. 1, 2). The Macolod Corridor is also with estimates ranging from 1.8 to 0.6 Ma (Sudo et al.
characterized by intense Pleistocene to Holocene volca- 2001). Volcanism continues through the present (Defant
nism. Sets of NE–SW and N–S trending faults and et al. 1989; Sudo et al. 2001), with the most recent
fractures are the dominant structural elements of the pyroclastic eruptions occurring at Taal Volcano
Macolod Corridor, but NW–SE trending structures approximately 5 ka (Listanco 1994; Martinez and
parallel to the arc are also present (Fig. 2; Förster et al. Williams 1999). The most recent eruption of lava from
1990; Pubellier et al. 2000; Aquino 2004). Taal was in 1977 (Miklius et al. 1991).
The crustal structure of the Luzon Arc is not well
known. Based on receiver functions in the vicinity of
Taal Volcano, which is located on the northern edge of Sampling
the Macolod Corridor, Besana et al. (1995) suggested
that the crust at this location is 34 km thick (Besana Silicic domes
et al. 1995). Their shear-wave velocity model is complex
with a thick, low-velocity zone at about 18 km, under- At least seven silicic domes are present near the western
lain by alternating low- and high-velocity zones. They and southeastern flanks of Mt. Makiling Volcano
interpret the 4 km thick low-velocity zone at 18 km due (Fig. 3). Six of these [Mt. Bijang, Calamba (Mt.
to partial melt and the alternating low- and high-velocity Mapinggoy), Turupiche (Tanauan Hill), San Antonio,
zones due to a complex lower crust. Olila and Bulalo] were sampled for this study. Bulalo
The origin of the Macolod Corridor is still a matter of and Olila Domes are near the site of the Mak-Ban
debate, and various hypotheses have been proposed. geothermal field, a commercial field, operated by Unocal
Defant et al. (1988) and many later workers have argued Philippines Incorporated (UPI). Drilling data from this
for a pull-apart rift zone, whereas Förster et al. (1990) field reveal that continuous hornblende dacite lavas
suggested a pull-apart structure due to a diffuse system similar to the Bulalo and Olila Dome compositions
of NW–SE oriented shears. Rifting due to transtensional underlie at least 10 km2 on the southeast flank of Mt.
motion has been proposed for the origin of the Macolod Makiling with a thickness varying from about 30 to
270

Fig. 3 Simplified geologic map of Makiling Volcano (modified from Oles 1991) showing the distribution of the domes around the volcano.
The domes are labeled as: 1 Bijang, 2 Calamba, 3 Turupiche (Tanauan Hill), 4 San Antonio, 5 Olila and 6 Bulalo

about 200 m. These units are buried by alluvium and Chemical analyses
younger pyroclastic deposits from Taal.
The least altered samples of the domes were For the pyroclastic deposits, pumice samples were ana-
obtained at recent road cuts or active construction lyzed. For the domes, where possible, the host rocks
sites (Mt. Bijang, Bulalo and Olila) and quarries were separated from the enclaves—enclaves that were
(Calamba). Where recent construction sites do not large enough were analyzed separately. When necessary
exist (Turupiche and San Antonio Domes), fresh samples were cleaned, passed through a chipmunk rock
samples were difficult to obtain due to deep tropical crusher and powdered using a ceramic flat plate grinder.
weathering. Fewer samples were obtained from these Three grams of this rock powder and 9.0 g of lithium
domes. tetraborate (Li2B4O7), along with 0.5 g of ammonium
nitrate (NH4NO3) (as an oxidizer) were fused in plati-
num crucibles at 1,000C for 20–30 min on an orbital
Ignimbrites mixing stage. The melt was then poured into platinum
molds making the glass disk that was analyzed using a
The Laguna de Bay Caldera produced voluminous Rigaku S-Max X-ray fluorescent (XRF) spectrograph.
pyroclastic flow deposits that were directed mostly to XRF major element analyses were reduced by a funda-
the northwest and north (Arpa et al. 1999a) and are mental parameter data reduction method using XRF-
well exposed in active quarries, construction sites and WIN software (Omni Instruments), while XRF trace
road cuts with 15 km of the margin of the caldera. element data were calculated using standard linear
Detailed descriptions of these pyroclastic deposits are regression techniques. All major and trace element
beyond the scope of this paper—based on our studies, whole rock analyses were done at the Michigan State
they are composed of at least three eruptive units that University. Major elements and Rb, Sr and Zr were
contain heterogeneous assemblages of pumice frag- analyzed using XRF. The rare earth elements, Nb, Ta,
ments. Samples of glassy pumice fragments were ob- Hf, Ba, Y, Th, U and Pb, were analyzed by laser
tained from 31 different sites that were distributed ablation inductively coupled plasma mass spectrometry
throughout the area. Based on crystal content, color on the same glass disks as used for XRF analyses. A
and vesiculation, all megascopically distinct pumice Cetac LSX200+ laser ablation system was used coupled
varieties were collected. with a Micromass Platform ICP-MS, using strontium,
 271

determined by XRF, as an internal standard. Trace represent averages of 2–4 individual analyses on differ-
element data reduction was done using MassLynx ent separates from the same sample. Oxygen was liber-
software. Element concentrations in the samples were ated from 1 to 3 mg of silicate and oxide by CO2 laser
calculated based on a linear regression method using fluorination and analyzed on a Finnigan MAT-251 gas-
well-characterized standards. Table 1 represents the source isotope-ratio mass spectrometer (Valley et al.
precision and accuracy of chemical analyses done in our 1995).
40
laboratory by replicating analyses of a standard (Jb-1a). Ar/39Ar ages were determined in samples from five
Representative chemical analyses of the domes are given domes (Mt. Bijang, Calamba, Tanauan Hill, Olila and
in Tables 2 and 3. Mt. Bulalo Domes) at the New Mexico Geochronolog-
Oxygen isotope data for mineral separates from ical Research Laboratory (Table 5). Hornblende sepa-
Calamba, Bulalo and Bijang Domes were collected using rates were analyzed by the incremental step heating
the CO2 laser fluorination/mass spectrometer at the method for all domes except Calamba, which contains
University of Wisconsin-Madison (Table 4). Values only trace amounts of hornblende. Because the Calamba

Table 1 JB-1a treated as an unknown for XRF and ICP-MS analyses: five different sample preparations

Variable N Mean Median SD SEa Givenb

XRF data
SiO2 5 52.05 52.16 0.273 0.122 52.41
TiO2 5 1.28 1.28 0.005 0.002 1.32
Al2O3 5 14.50 14.5 0.036 0.016 14.53
Fe2O3(t) 5 9.00 9 0.005 0.002 9.05
MnO 5 0.14 0.14 0.000 0.000 0.15
MgO 5 7.82 7.81 0.069 0.031 7.83
CaO 5 9.29 9.29 0.008 0.004 9.31
Na2O 5 2.71 2.71 0.023 0.010 2.73
K2O 5 1.38 1.38 0.005 0.002 1.40
P2O5 5 0.27 0.27 0.004 0.002 0.26
Total 5 98.45 98.49 0.171 0.076 98.99
Cr 5 373 372 6 3 392
Ni 5 138 138 2 1 139
Cu 5 74 75 6 3 57
Zn 5 74 74 1 0 82
Rb 5 40 41 2 1 39
Sr 5 460 460 2 1 442
Y 5 23 23 1 0 24
Zr 5 139 139 1 0 144
Nb 5 22 22 1 0 27
La 5 27 29 6 3 38
Ba 5 455 461 22 10 504
ICP-MS data
Zr 5 147.13 147.33 1.97 0.88 144
Ba 5 473.72 475.75 5.74 2.57 504
La 5 38.24 38.32 0.48 0.22 37.6
Ce 5 62.63 63.16 1.25 0.56 65.9
Pr 5 7.29 7.39 0.25 0.11 7.3
Nd 5 27.36 27.57 0.82 0.36 26.9
Sm 5 5.47 5.52 0.20 0.09 5.07
Eu 5 1.54 1.55 0.07 0.03 1.46
Gd 5 5.08 5.13 0.14 0.06 4.67
Tb 5 0.79 0.8 0.02 0.01 0.69
Y 5 24.57 24.55 0.17 0.08 24
Dy 5 4.25 4.28 0.09 0.04 3.99
Ho 5 0.83 0.84 0.03 0.01 0.71
Er 5 2.30 2.31 0.07 0.03 2.18
Yb 5 2.27 2.29 0.09 0.04 2.1
Lu 5 0.33 0.34 0.01 0.01 0.33
V 5 186.88 186.39 3.49 1.56 205
Cr 5 380.92 377.91 8.72 3.90 392
Nb 5 27.21 26.98 0.67 0.30 26.9
Hf 5 3.73 3.78 0.18 0.08 6.65
Ta 5 1.83 1.86 0.08 0.04 1.93
Pb 5 5.85 5.92 0.18 0.08 6.76
Th 5 10.02 10.11 0.36 0.16 9.03
U 5 1.64 1.67 0.07 0.03 1.67
a
Standard error = standard deviation divided by the square root of the number of samples
b
http://www.aist.go.jp/RIODB/geostand/igneous.html
272

Table 2 Representative major element analyses (in wt%) of hosts and enclaves from domes

Sample Dome Lon(N) Lat(E) SiO2 TiO2 Al2O3 Fe2O3(t) MnO MgO CaO Na2O K2 O P2O5 Totals

Sample Calamba 14.144 121.162 53.28 0.81 18.11 7.85 0.15 4.15 8.95 3.37 1.48 0.16 98.31
17-2a Calamba 65.10 0.65 16.03 4.41 0.10 1.33 3.76 4.28 2.97 0.16 98.79
17-2b Calamba 50.96 0.89 17.99 9.10 0.15 5.24 10.38 3.00 0.90 0.14 98.75
17-2c Calamba 52.44 0.97 17.67 9.08 0.16 4.43 8.99 3.31 1.23 0.16 98.44
17-2e Calamba 65.51 0.66 15.98 4.47 0.10 1.28 3.65 4.27 3.01 0.17 99.10
17-2f Calamba 64.61 0.64 16.23 4.35 0.10 1.34 3.76 4.29 2.94 0.16 98.42
17-2g Calamba 66.19 0.64 16.07 4.36 0.10 1.31 3.81 4.22 2.97 0.16 99.83
17-2h Calamba 64.64 0.69 15.90 4.74 0.11 1.39 3.80 4.23 2.93 0.16 98.59
17-2i Calamba 65.58 0.66 15.81 4.46 0.11 1.43 3.52 4.22 3.16 0.16 99.11
15-3B Mt. Bijang 14.163 121.166 69.10 0.37 14.56 2.94 0.08 0.88 2.73 3.66 3.45 0.10 97.87
18-1a Mt. Bijang 68.60 0.37 14.77 2.93 0.08 0.91 2.68 3.60 3.49 0.10 97.53
18-1TAV-a Mt. Bijang 67.40 0.38 15.08 3.05 0.08 1.00 2.89 3.36 3.74 0.10 97.08
18-1TAV-b Mt. Bijang 68.92 0.36 14.65 2.75 0.08 0.88 2.64 3.63 3.44 0.10 97.45
18-1TAV-c Mt. Bijang 67.89 0.38 14.85 2.90 0.08 0.92 2.67 3.68 3.49 0.10 96.96
18-1f Mt. Bijang 68.82 0.37 14.53 2.91 0.08 0.89 2.62 3.46 3.86 0.10 97.64
18-1g Mt. Bijang 68.07 0.40 14.63 3.23 0.09 0.98 2.71 3.61 3.51 0.10 97.33
18-1h Mt. Bijang 69.00 0.39 14.66 3.11 0.08 0.91 2.60 3.55 3.69 0.11 98.10
15-4A Mt. Bijang 69.03 0.37 14.69 2.78 0.08 0.92 2.68 3.67 3.54 0.11 97.87
15-4B Tanauan Hill 14.091 121.129 67.08 0.41 15.04 3.27 0.09 1.01 2.92 3.85 3.46 0.12 97.25
18-2a Tanauan Hill 69.13 0.38 14.59 2.99 0.09 0.88 2.69 3.71 3.61 0.11 98.18
18-2b Tanauan Hill 68.07 0.42 14.90 3.49 0.10 1.02 3.01 3.78 3.40 0.12 98.31
18-2c Tanauan Hill 68.32 0.37 14.61 2.81 0.09 0.80 2.47 3.87 3.51 0.11 96.96
20-1a Tanauan Hill 68.62 0.38 14.85 2.88 0.09 0.85 2.65 3.92 3.39 0.11 97.74
20-1b Tanauan Hill 68.43 0.38 14.70 2.91 0.09 0.85 2.57 3.79 3.62 0.11 97.45
20-1c Mt. Bulalo 14.110 121.209 69.78 0.37 14.37 2.84 0.09 0.80 2.39 3.76 3.60 0.11 98.11
22-50TAV Mt. Bulalo 49.87 1.06 17.82 10.66 0.18 4.32 9.67 3.10 1.02 0.21 97.91
22-51TAV Mt. Bulalo 50.71 1.03 17.75 10.20 0.18 4.19 9.27 3.11 1.14 0.21 97.79
22-51tavE1 Mt. Bulalo 51.28 1.04 17.69 10.31 0.18 4.03 9.29 3.13 1.16 0.21 98.32
22-51tavE2 Mt. Bulalo 68.54 0.40 15.62 3.18 0.10 1.02 2.98 4.32 3.03 0.12 99.31
15-1A Mt. Bulalo 68.95 0.40 15.37 3.11 0.10 1.05 2.95 4.32 3.07 0.13 99.45
15-1B Mt. Olila 14.082 121.211 52.52 0.85 17.65 9.29 0.17 4.36 8.83 2.97 1.29 0.21 98.14
22-52TAV Mt. Olila 70.04 0.38 14.99 2.88 0.09 0.88 2.78 4.00 3.20 0.13 99.37
22-53TAV Mt. Olila 68.82 0.39 15.14 2.88 0.09 0.90 2.69 4.14 3.19 0.12 98.36
22-54TAV Mt. Olila 69.64 0.37 14.84 2.80 0.09 0.86 2.64 4.07 3.24 0.11 98.66
22-55TAV Mt. Olila 69.09 0.38 15.24 2.90 0.09 0.90 2.92 4.12 3.16 0.13 98.93

Dome rock lacked hornblende, plagioclase was ana- are the mafic phases. Samples taken from surface
lyzed. The methods used can be found at http:// exposures of domes are glassy and microlitic in all the
www.geoinfo.nmt.edu/publications/openfile/argon/ domes except Calamba, where some samples are devit-
home.html. Because of the young age of the samples, rified. Mafic enclaves are found in most domes, but none
large amounts of hornblende (
50–115 mg) were ana- were observed in the Bijang Dome. Mafic enclaves vary
lyzed. Two samples, one from Bulalo Dome and one in size from about 1 to 25 cm. Based on the limited
from Olila Dome, were replicated four times each to outcrops, the abundance of these enclaves is generally
assess age homogeneity and to improve overall preci- about 5%, but locally it is as much as 20% in some
sion. Typically, 6–8 heating steps were used to construct quarry exposures in the Calamba Dome. The major
the age spectra (data are available from the corre- differences between the large and small enclaves are the
sponding author upon request). size of the grains in the enclaves and the textural evi-
Minerals were analyzed using a Cameca SX 100 dence of interaction with the host magma. Some large
EMPA at the University of Michigan. In general, an enclaves exhibit crenulate chill margins and vesicular
accelerating potential of 15 kV was used. Other settings interiors, whereas small enclaves have interacted with
such as beam spot size, counting time, and current the host magma.
varied based on the phase being analyzed. The domes, except for Calamba Dome, contain pla-
gioclase, hornblende and orthopyroxene in the order of
decreasing abundance, along with some trace phases.
Texture and mineralogy of the domes For Calamba Dome, the order of abundance is plagio-
clase, orthopyroxene, clinopyroxene, with minor
The mineral assemblages of the domes are very similar amounts of hornblende. Plagioclase compositions in the
with the exception of Calamba. All dome samples domes are similar. Anorthite content decreases towards
contain 15–25% phenocrysts. The dominant phase is the edges of the grains with the edge composition
plagioclase followed by hornblende, except in the ranging from An31 to An41. Plagioclase cores generally
Calamba Dome where orthopyroxene and clinopyroxene contain a maximum anorthite content of An55. Centers
Table 3 Representative trace element analyses (in ppm) of hosts and enclaves from domes

Sample Dome Rb Sr Zr V Cr Y Nb Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu Hf Ta Pb Th U

17-2a Calamba 36 522 83 245 45 19 5 357 15 29 3.6 14.6 3.4 1.2 3.4 0.5 3.1 0.7 1.9 2.0 0.3 2.1 0.3 8 4.6 1.0
17-2b Calamba 74 341 185 103 6 25 9 650 24 49 5.6 20.9 4.5 1.2 4.4 0.7 3.9 0.8 2.5 2.9 0.4 4.5 0.5 12 10.3 3.1
17-2c Calamba 20 535 61 277 64 18 4 263 11 22 2.8 11.8 2.9 1.0 3.1 0.5 2.9 0.6 1.7 1.8 0.3 1.6 0.2 6 3.0 0.7
17-2e Calamba 26 523 80 277 17 21 5 304 13 25 3.2 13.6 3.2 1.1 3.4 0.5 3.2 0.7 2.1 2.1 0.3 2.2 0.3 8 4.0 0.8
17-2f Calamba 69 335 196 103 5 26 10 666 24 50 5.8 21.7 4.7 1.2 4.5 0.7 4.0 0.9 2.5 3.0 0.4 4.6 0.5 14 10.1 3.2
17-2g Calamba 72 352 193 105 6 24 10 672 24 49 5.6 21.2 4.5 1.3 4.5 0.7 4.3 0.9 2.6 2.9 0.4 4.8 0.7 13 11.4 4.3
17-2h Calamba 75 344 191 105 5 23 10 670 23 49 5.5 20.3 4.3 1.2 4.3 0.7 3.6 0.8 2.2 2.5 0.4 4.4 0.6 17 11.0 4.3
17-2i Calamba 73 342 192 121 6 23 11 667 24 50 5.8 21.4 4.6 1.3 4.6 0.7 3.8 0.8 2.4 2.8 0.4 4.7 0.6 18 11.0 4.3
15-3B Calamba 80 319 203 125 8 25 11 693 23 52 5.5 20.6 9.6 1.2 4.3 0.7 3.9 0.9 2.6 2.7 0.4 4.3 0.6 23 10.5 3.5
18-1a Mt. Bijang 95 307 160 70 6 13 9 780 23 42 4.1 13.4 2.7 0.7 2.6 0.4 2.1 0.4 1.3 1.8 0.3 3.8 0.6 35 15.3 6.0
18-1TAV-a Mt. Bijang 93 309 168 69 7 12 9 779 22 41 4.1 13.6 2.7 0.8 2.5 0.3 2.1 0.4 1.3 1.6 0.3 3.9 0.7 22 14.1 6.0
18-1TAV-b Mt. Bijang 90 327 164 71 6 13 9 762 23 42 4.3 14.8 2.8 0.8 2.7 0.4 2.1 0.5 1.4 1.9 0.3 4.1 0.7 37 16.5 5.7
18-1TAV-c Mt. Bijang 96 302 163 60 6 13 8 797 23 42 4.2 13.8 2.5 0.8 2.5 0.4 1.9 0.4 1.3 1.7 0.2 4.0 0.6 36 15.8 5.8
18-1f Mt. Bijang 89 315 170 68 7 12 9 766 22 41 4.1 13.5 2.4 0.8 2.5 0.4 2.0 0.5 1.4 1.7 0.3 4.1 0.7 24 15.2 6.0
18-1g Mt. Bijang 96 297 159 65 5 12 9 776 22 41 4.0 13.2 2.5 0.8 2.6 0.4 1.9 0.5 1.3 1.6 0.3 3.9 0.7 35 15.7 5.8
18-1h Mt. Bijang 92 305 162 81 6 13 9 776 23 43 4.3 14.4 2.9 0.8 2.8 0.4 2.1 0.5 1.5 1.8 0.3 4.1 0.7 36 16.1 6.1
15-4A Mt. Bijang 107 304 167 76 7 12 9 817 22 45 4.2 13.4 2.5 0.8 2.6 0.4 2.0 0.5 1.5 1.6 0.2 3.8 0.7 33 15.9 4.6
15-4B Mt. Bijang 99 310 163 64 6 13 9 814 22 45 4.3 13.8 2.6 0.8 2.7 0.5 2.1 0.5 1.5 1.7 0.2 3.6 0.7 33 15.5 4.5
18-2a Tanauan Hill 87 325 159 71 4 14 8 772 23 42 4.3 14.0 2.8 0.9 2.8 0.4 2.1 0.5 1.4 1.7 0.3 3.6 0.5 31 14.2 5.3
18-2b Tanauan Hill 93 299 161 908 998 46 2 19 2 5 1.0 6.8 3.1 1.5 5.0 1.0 7.4 1.7 5.0 4.9 0.8 1.9 0.2 9 0.1 0.1
18-2c Tanauan Hill 81 324 157 76 5 14 8 756 23 42 4.2 14.3 2.6 0.9 2.7 0.4 2.2 0.5 1.5 1.9 0.3 3.7 0.6 29 14.3 5.2
20-1a Tanauan Hill 95 287 169 60 4 12 7 676 20 38 3.8 13.0 2.6 0.8 2.6 0.4 1.9 0.4 1.3 1.8 0.3 3.4 0.5 29 13.0 4.7
20-1b Tanauan Hill 95 306 163 58 5 14 8 801 23 43 4.2 13.8 2.5 0.8 2.6 0.4 1.9 0.4 1.3 1.9 0.3 3.6 0.6 27 13.5 4.4
20-1c Tanauan Hill 96 299 167 63 4 12 7 671 20 37 3.7 12.7 2.4 0.8 2.5 0.4 2.0 0.5 1.4 1.5 0.2 3.3 0.5 28 12.9 4.8
22-50TAV Mt. Bulalo 90 269 174 185 49 14 3 149 7 13 1.7 7.5 1.9 0.7 2.1 0.4 2.4 0.5 1.4 1.3 0.2 1.4 0.2 6 2.2 0.7
22-51TAV Mt. Bulalo 23 597 82 125 11 29 18 1,828 53 91 9.3 31.4 5.4 1.7 5.7 0.8 5.2 1.0 3.2 3.8 0.7 10.2 1.8 49 35.6 12.2
22-51tavE1 Mt. Bulalo 31 584 89 352 5 22 3 317 14 27 3.7 16.8 3.9 1.3 4.1 0.6 3.9 0.8 2.2 2.2 0.3 2.3 0.2 6 3.8 0.9
22-51tavE2 Mt. Bulalo 28 586 86 348 6 23 4 353 15 28 4.0 17.3 4.0 1.4 4.4 0.7 4.0 0.8 2.3 2.2 0.4 2.7 0.3 6 4.5 1.3
15-1A Mt. Bulalo 81 339 170 79 9 13 10 843 23 47 4.4 14.3 2.8 1.0 2.8 0.5 2.1 0.5 1.5 1.7 0.3 3.5 0.7 34 14.1 4.8
15-1B Mt. Bulalo 85 332 168 65 7 14 8 841 23 45 4.3 14.7 2.9 0.9 2.9 0.5 2.5 0.6 1.7 2.0 0.3 4.0 0.7 34 16.7 4.4
22-52TAV Mt. Olila 31 504 87 287 3 20 3 291 13 24 3.3 15.0 3.6 1.2 3.7 0.6 3.5 0.7 1.9 1.9 0.3 2.2 0.2 7 3.7 0.9
22-53TAV Mt. Olila 86 310 181 180 13 12 2 211 8 15 2.0 8.4 2.0 0.7 2.1 0.3 2.2 0.4 1.2 1.2 0.2 1.5 0.2 5 3.0 0.8
22-54TAV Mt. Olila 86 307 183 54 5 13 8 794 24 42 4.3 15.2 2.6 0.9 2.7 0.4 2.5 0.5 1.5 1.8 0.3 4.2 0.7 22 14.8 5.1
22-55TAV Mt. Olila 86 295 173 57 5 13 8 802 23 41 4.1 13.9 2.5 0.8 2.6 0.4 2.4 0.5 1.4 1.6 0.3 4.3 0.7 23 14.5 5.2
22-56TAV Mt. Olila 80 323 165 60 6 14 9 885 26 45 4.6 15.8 2.9 0.9 2.8 0.4 2.6 0.5 1.7 2.0 0.4 4.6 0.8 25 16.5 5.9
273
274

Table 4 Data for oxygen isotope calculations of the domes. Oxygen isotope data presented in d notation (&) relative to V-SMOW

Bulalo Calamba Bijang

SiO2 (wt%) 68.5 65.5 68.5

Magnetite d 2.36±0.08 2.83 (1 analysis) 2.48±0.12
Hornblende d 4.90±0.03 4.82±0.14 4.67±0.06
Difference d 2.54 1.99 2.19
Calculated whole rock d 5.90 6.29 6.06
T (clinopyroxenemagnetite)a 908C 1,060C 999C
D(meltamphibole)b 1.96 1.28 1.96
Melt d18O&c 6.36 6.1 6.13

Values are averages for duplicate analyses. Errors are one standard deviation. See text for discussion
a
Calculated using the A factor from Chiba et al. (1989)
b
D(meltamphibole)=0.61(SiO2 wt%)2.72 (Bindeman et al. 2004)
c 18
d O& melt=d18O& Cpx + D(meltamphibole) (Bindeman et al. 2004)

of some plagioclase grains are as high as An84. paper. The silicic samples contain only orthopyroxene
Some plagioclase grains contain reverse zoning. Ortho- whereas the more mafic samples contain both orthopy-
pyroxenes in all domes are similar in composition roxene and clinopyroxene. Magnetite is the only oxide
(Wo1.3–3.9En68.4–70.6Fs27.5–29.9). Hornblendes compose and trace amounts of apatite and zircon are present.
25–30% of the phenocrysts assemblage and have little The mineral compositions are similar to those ob-
compositional variation among domes (typical horn- served in the domes. Plagioclase rim compositions are
blende formula of ðNa0:07 K0:43 ÞCa1:68 ðMg3:28 Mn0:08 similar to the plagioclase from the domes (An35). Clin-
Fe0:82 Ti0:22 Fe0:56 Al0:37 Þ½Si6:13 Al1:87 O22 (OH)2 ). The Ca- opyroxene compositions are Wo38–45En41–46Fs14–15.
lamba Dome contains both orthopyroxene and Orthopyroxene compositions are Wo3–4En66–70Fs26–30.
clinopyroxene (Wo40.7–43.4En43.2–45.3Fs13.0–15.0) as the
dominant mafic phase in subequal amounts. The most
abundant accessory phases in the domes are Fe–Ti Chemistry of the domes compared to Makiling lavas
oxides; only magnetite occurs in most domes, while both and Laguna de Bay ignimbrites
magnetite and ilmentite occurs in Calamba Dome.
Zircon and apatite occur as trace phases in all domes. Representative chemical analyses for the domes are gi-
All the phases are unaltered. ven in Tables 2 and 3. A complete data file is given as an
electronic supplement. With the exception of Calamba,
the domes are dominated by compositions straddling
Mineralogy of the Laguna de Bay pyroclastic deposits from the dacitic to rhyolitic transition. Glassy matrix is
present at all domes, indicating that the rapidly quen-
Pumice clasts within the ignimbrites from the Laguna de ched dome carapace has not been removed by erosion.
Bay Caldera are nearly aphyric, containing less than 10% Mafic enclaves are present in three of the domes: Olila,
phenocrysts. The dominant phases are plagioclase, Bulalo and Calamba. The host compositions of all the
orthopyroxene and clinopyroxene. Hornblende occurs in domes are similar except for Calamba. The host com-
some of the silicic samples. Details of the variations of the positions of these domes range from 67.1 to 71.1 wt%
mineralogy among the ignimbrite units are beyond the SiO2, with a mean of 70.0 wt% (SD 0.77), whereas the
scope of this paper and will be presented in a subsequent host composition of the Calamba Dome ranges from
40
Table 5 Summary of Ar/39Ar
hornblende results Sample Irrad N K/Ca Age (ka) ±1r Eruption age ±1r

Bulalo
BD-1A-a NM-155 5 0.052 0.035 0.020
BD-1A-b NM-155 2 0.052 0.014 0.013
BD-1A-c NM-155 5 0.076 0.015 0.013
BD-1A-d NM-155 6 0.084 0.005 0.014 15±7 ka
Bulalo
BD-1B NM-155 6 0.148 0.046 0.009 46±9 ka
Olila
020522-56 TVA-a NM-163 4 0.053 0.030 0.015
020522-56 TVA-b NM-163 4 0.052 -0.009 0.019
020522-56 TVA-c NM-163 3 0.053 0.024 0.014
020522-56 TVA-d NM-163 3 0.052 -0.042 0.024 11±8 ka
Bijang
020518-1g NM-163 2 0.046 0.066 0.014 66±14a
N number of steps defining Tanauan Hill
plateau 020520-1c NM-163 4 0.051 0.043 0.017 43±17
a
Maximum age
 275

Fig. 4 Alkali–silica
classification diagram (LeBas
et al. 1986) showing the
composition of the domes
compared to the lavas of
Makiling Volcano. Dashed line
outlines Laguna de Bay
samples; solid line outlines
Makiling Volcano samples.
Empty symbols are enclaves

65.3 to 66.3 wt% SiO2, with a mean of 65.8 wt% (SD et al. 1986; Fig. 4). Examples of the chemical variation
0.32). Mafic enclaves have similar compositions in all of the domes compared to Makiling lavas and pumice
domes where they occur and range from 50.9 to fragments from Laguna de Bay ignimbrites are shown in
54.2 wt% with a mean of 52.2 wt% (SD 1.18). The host Figs. 4, 5, 6 and 7.
samples plot in the dacite–rhyolite fields and the mafic Chemical analyses of Makiling Volcano are from an
enclaves plot in the basalt–basaltic andesite fields (LeBas unpublished masters thesis of J.B. Cruz, unpublished

Fig. 5 Chemical comparison of the domes with Makiling lavas and dome samples and silicic lava samples; the Laguna de Bay samples
pumice samples from Laguna de Bay ignimbrites. a MgO (wt%), are different with respect to MgO, Zr and Fe2O3. Note that there is
b Zr (ppm), c Fe2O3 (wt%) and d Sr (ppm) versus SiO2 (wt%). a large variation in Sr concentrations; the highest SiO2 samples
Dashed line outlines Laguna de Bay samples; solid line outlines have relatively high concentrations of Sr (>200 ppm)
Makiling Volcano samples. Note the general similarities of the
276

Fig. 6 Chondrite-normalized (Sun and McDonough 1989) REE 45–55 wt% SiO2 (outlined in a solid line). Note the similarity in the
diagram showing a variation in the Calamba Dome, b variation in silicic lavas and host samples for the dome, particularly the
all other domes, c variation in the Makiling lavas (data points) and Calamba Dome and the similarity of mafic samples from Makiling
Laguna de Bay ignimbrites with greater than 65 wt% SiO2 and the mafic enclaves. Although there is some overlap, the Laguna
(outlined in a dashed line), d variation of Makiling lavas with de Bay silicic samples are distinctly higher in all REEs

Unocal data and unpublished analyses done at the any Eu anomaly precludes any significant plagioclase
MSU (available as an electronic supplement). Chemical fractionation.
analyses of the Laguna de Bay ignimbrites are from The REE distribution of the host lavas from the
unpublished analyses done at the MSU (available as an domes is similar except for the Calamba Dome (Fig. 6a).
electronic supplement). Chemical variations within the All are characterized by a well-developed concave up-
mafic enclaves and silicic host of the domes overlap ward trend in the middle REEs (Fig. 6a, b), a middle
with the chemical variations within the Makiling lavas. rare earth depletion pattern. This pattern is subdued in
This is best displayed in Figs. 4 and 5, which show the Calamba Dome. High silica lavas from Makiling
alkalis, MgO, Fe2O3, Zr and Sr variations with SiO2. contain a few samples that have a slight concave upward
The composition of the mafic enclaves and host com- pattern (Fig. 6c), but most do not. Most of the Laguna
positions of the domes are very similar to the compo- de Bay silicic samples contain distinctly higher concen-
sitions of the Makiling lavas. However, the trations of REEs (Fig. 6c). The mafic enclaves from the
compositions of pumice fragments from Laguna de Bay domes are similar to the mafic portions of the Makiling
ignimbrites are different from the domes. This is best lavas (Fig. 6d).
documented in Figs. 5 and 6, which show distinct dif- Spider diagrams normalized to primitive mantle (Sun
ferences in MgO, Fe2O3, Zr, Sr and the rare earth and McDonough 1989) show that all silicic samples are
elements (REEs) in the domes and ignimbrites. Also, depleted in Nb and Ti, the characteristic subduction
the domes and Makiling lavas contain similar trends, signature (Fig. 7). The domes and Makiling silicic lava
except for the much larger variation in Sr in the samples have very similar patterns (Fig. 7a–c), whereas
intermediate compositions in the Makiling lavas the silicic Laguna de Bay ignimbrites have distinctly
(Fig. 5c). The most striking aspect of the Sr variation is higher values of all elements on these diagrams. The
the high concentrations for relatively high SiO2 values enclaves from the domes are similar to the more mafic
(>200 ppm). The high Sr concentrations and lack of Makiling lavas (Fig. 7d).
 277

Fig. 7 Primitive mantle-normalized (Sun and McDonough 1989) REE diagrams showing a Calamba Dome; b all other domes; c silicic
(>65 wt% silica) for Makiling lavas (data points) and Laguna de Bay ignimbrites (shaded area) and d Makiling lavas with 45–55 wt% silica

Age constraints relatively well-preserved morphology. The age of the

Makiling lavas is poorly constrained (J.B. Cruz,
40
Ar/39Ar ages were determined for five domes: Mt. unpublished thesis). Dates of 510–180 ka have recently
Bijang, Calamba, Tanauan Hill, Olila and Mt. Bulalo been reported (Sudo et al. 2001). However, unpublished
(Table 5). All are plateau ages. Two samples had four U/Pb SHRIMP analyses of zircons from cores recovered
replicates each and yielded internally reproducible re- from the Mak-Ban geothermal field southeast of Mak-
sults. The Bulalo Dome has an eruptive age of 15±7 ka iling provide evidence that the entire edifice is signifi-
and the Olila Dome has an eruptive age of 11±8 ka. cantly younger than 350 ka. Two thick rhyolitic ash flow
These domes are adjacent to one another (Fig. 3) and tuff sequences intersect by drilling at depths of about 1
are considered identical in age. For Tanauan Hill and and 1.5 km yielded U/Pb zircon SHRIMP ages of 350
Mt. Bijang, only one 40Ar/39Ar age spectrum from and 500 ka, respectively (J. Lowenstern, personal com-
hornblende was determined, and we are less confident munication). These units are not exposed at the surface.
that the ages truly represent the eruption age. The However, their major element chemistry is similar but
apparent eruption age for Tanauan Hill is 43±17 ka more evolved than surface dacite lavas and domes.
and Mt. Bijang is 66±14 ka. The eruption age defined Considering their stratigraphic position and thickness
by plagioclase for the Calamba Dome is difficult to (locally >200 m), these silicic ash flows probably rep-
determine because the age spectrum contains two pla- resent caldera-related deposits that predate the Makiling
teaus. The first plateau is consistent with an 80 ka age, episode of volcanism (Dimabuyu et al. 2004).
whereas the second plateau is consistent with a 250 ka The Laguna de Bay ignimbrites have reported
age. Either could represent the eruption age of the unit, radiocarbon ages from 27 to 50 ka (Catane and Arpa
but the plateaus could also result from excess argon (the 1998; Arpa et al. 1999a; Catane et al. 2004). However,
250 ka plateau) or argon gain (the 80 ka plateau). the upper limit of these ages is poorly constrained due to
All the domes are relatively young, with the possible the limits of radiocarbon dating.
exception of Calamba. Both Bulalo and Olila Domes are
between 22 and 3 ka with good age constraints. These
ages are consistent with the well-preserved morphology Oxygen isotope data
of the domes, their glassy matrix texture and the lack of
any overlying units. The ages of other domes are not as Data were collected from magnetite and hornblende
well constrained, but they most likely erupted <250 ka grains from Calamba, Bijang and Bulalo Domes in order
based on the available age information and their to determine the whole rock d18O values for each dome
278

and less than 0.5& for Calamba. The hornblende d18O

values were also very similar, with a difference of less
than 0.25& among all three domes. Whole rock d18O
values for the domes were calculated assuming equilib-
rium with the phenocrysts using an assumed tempera-
ture of 950C (Clayton et al. 1989). The calculated whole
rock d18O values for the three domes were: Bulalo 5.9&,
Calamba 6.29& and Bijang 6.06&. These values are
slightly higher than the range of oceanic arcs (4.9–5.8&)
(Eiler et al. 2000). The similarities of the d18O values of
the mineral phases and the whole rock indicate that the
minerals used in the calculations were in equilibrium
with each other. As a check on these calculations we
calculated the temperature based on amphibole–mag-
netite equilibrium using the experiments from Chiba
et al. (1989) (Table 4). The fractionation factors for
amphibole can be assumed to be similar to clinopyrox-
ene (Hoefs 2004). Using the A factor from Chiba et al.
(1989) the d18O of the melt can be calculated from
d18 O melt = d18 O cpx + ð0.061  SiO2 wt% minus; 2.72Þ
(Bindeman et al. 2004). The range from 6.10 to 6.36& is
similar to the calculations assuming a 950C tempera-
ture (Table 3). This is important because it indicates that
the d18O oxygen values from the minerals have not been
reset and therefore reflect the values of the source rock
(Bindeman and Valley 2003). These values preclude a
derivation from melting a hydrothermally altered crust
and are consistent with a derivation from a mantle-
equilibrated source rock either by partial melting or
fractional crystallization.
Fig. 8 a K2O/Na2O versus K2O + Na2O plots showing the
composition of the silicic samples from the domes, Makiling lavas
and Laguna de Bay ignimbrites compared to experimental melt Discussion
compositions from melting of primitive basalts and andesites. Open
triangle with dashed line is a starting composition of an Mg-rich
andesite at 0.1 MPa; closed triangle with dotted line is a starting The occurrence of abundant silicic rocks in intra-oceanic
composition of an Mg-rich andesite at 200 MPa; open squares are arc settings has only recently been appreciated (Fiske
starting compositions of Medicine Lake basalts at 200 MPa (Grove et al. 2001; Tamura and Tatsumi 2002; Leat et al. 2003;
et al. 2003). Closed squares and closed diamonds are starting
compositions of primitive basalts from Mt. Shasta with 3.8 and 5.0
Smith et al. 2003; Tamura et al. 2003; Vogel et al. 2004,
wt% H2O, respectively (Müntener et al. 2001). The pluses with a 2006). Previously, it was believed that the low abun-
line drawn through them is a starting composition of a tholeiitic dance of silicic rocks in these settings was due to the
basalt with extreme fractional crystallization/partial melting with absence of continental crust. All the above investigators
about 3.5% liquid remaining at the highest K2O/Na2O values have attributed the origin of these silicic magmas to the
(Villiger et al. 2004). Note that high K2O/Na2O samples are not
produced by fractional crystallization or partial melting of partial melting of crustal rocks with compositions
primitive basalt or Mg-rich andesites. b K2O/Na2O versus ranging from andesite to amphibolite.
K2O + Na2O plots showing the composition of the samples from Various workers (Beard and Lofgren 1991; Müntener
the domes, Makiling lavas and Laguna de Bay ignimbrites (shaded et al. 2001; Grove et al. 2003; Villiger et al. 2004) have
areas) with less than 55 wt% SiO2 compared to experimental data
from Sisson et al. (2005) for melting high-K basalts. Solid symbols shown that low-K basaltic or low-K andesitic material
are the high-K basaltic starting materials; open symbols are partial cannot be a source for these high K2O/Na2O calc-
melts produced under various conditions and melt fractions (12– alkaline magmas (Fig. 8a). Others have hypothesized
34%). Lines connect the same starting compositions with different that the existence of an evolved crustal component is
experimental conditions. Percent liquid in the experiments varied
from 34 to 12%. Refer a for the compositions of the silicic samples
needed to explain the petrogenesis of these high K2O/
Na2O calc-alkaline magmas (Patiño Douce and
McCarthy 1998; Patiño Douce 1999). However, Sisson
(Table 4). Minerals were used rather than whole rock et al. (2005), using more evolved basaltic starting com-
because of the secondary hydration of the glass. The positions, have recently demonstrated that partial
d18O values collected from each mineral type were sim- melting or advanced fractional crystallization of med-
ilar among the three domes. The magnetite d18O values ium- to high-K basaltic rocks can produce liquids with
varied by less than 0.12& between Bulalo and Bijang high K2O/Na2O ratios (Fig. 8b).
 279

All the calc-alkaline silicic rocks in the Macolod cold crust. Remelting of hot, stalled mostly crystalline
Corridor contain high alkalis and high K2O/Na2O ratios magmas is an energy efficient process to produce magma
(Fig. 8a). In Fig. 8a, the silicic samples (>65 wt% SiO2) (Tamura et al. 2003). Heat for melting is transferred
from the domes, Makiling lavas and Laguna de Bay from the emplacement of other mantle-derived magmas
ignimbrites are compared to experimental data for to hot plutons. An alternative process is by the extrac-
crystal fractionation for primitive basalts and primitive tion of residual melt from stalled, partially crystallized
andesites (Müntener et al. 2001; Gaetani et al. 2003; magma (Bachmann and Bergantz 2003, 2004). Each
Grove et al. 2003; Villiger et al. 2004). The plus symbols process would produce new magmas of similar compo-
in Fig. 8a represent experimental data from nearly per- sitions.
fect fractionation of tholeiitic basalt (the last liquid is Few Nd and Sr isotopic studies have been published
less than 3.5% of the original material) with liquids for igneous rocks in the Macolod Corridor; however,
approaching dacite in composition (Villiger et al. 2004). there are some relevant unpublished data. Isotopic
A conclusion from these experiments is that without an studies on Taal lavas and Laguna de Bay pyroclastic
evolved source, it is difficult to produce liquids with high deposits showed relatively depleted ratios for both
alkalis and high K2O/Na2O ratios by the partial melting volcanic systems (Mukasa et al. 1994a). For Taal
of Mg-rich andesites or primitive basalts. samples, Sr isotopic values range from 0.70443 to
Recently, Sisson et al. (2005), using evolved basalts in 0.70472 and the Nd isotopic values range from 0.51279
melting experiments, produced melts that have very to 051285, whereas these ratios in Laguna de Bay
similar alkalis and K2O/Na2O values observed for the samples range from 0.70409 to 70426 and 0.51276 to
silicic rocks in Macolod Corridor (Fig. 8b). One of their 0.51285, respectively (Mukasa et al. 1994b). The
143
starting compositions (Yos-55a) is a gabbro that has Nd/144Nd versus 87Sr/86Sr diagram shown in
alkali concentrations similar to some of the mafic sam- Mukasa et al.’s (1994b) study demonstrates a close
ples in the Laguna de Bay ignimbrite and Makiling cluster of data points similar to the Izu-Honshu Arc of
basaltic lavas (Fig. 8b). Their experiments demonstrate Japan falling well above the bulk earth values. No
that the evolved liquids produced by partial melting of evidence of evolved continental crust can be observed
medium to high-K basaltic magma or extreme crystal in these data. Mukasa provided one analysis of a silicic
fractionation of moderate to high-K basalt could sample from Makiling with an Sr ratio of 0.70417 and
produce the high K2O/Na2O silicic magmas present in Nd ratio of 0.51286 (S.B. Mukasa, personal commu-
the Macolod Corridor. More experiments on high-K nication). Seven unpublished Sr isotope analyses of
andesites are needed. Cristo-Banahaw and five from the Makiling area,
Our preferred model for the generation of these high- including Olila, Bijang and Mabilog Domes (unpub-
silica magmas is by partial melting of calc-alkaline, lished data, Unocal Philippines, Inc.; analyses done at
mantle-derived, moderate to K-rich crystallized evolved the UC Santa Barbara by J. Mattinson) were made
magmas that have ponded and crystallized in the low- to available. The Cristo-Banahaw data range from
mid-crust. However, partial melting of the lower crustal 0.70409 to 0.70449, whereas the Makiling data range
mid- to high-K mafic rocks could produce these silicic from 0.70430 to 0.70443. Olila is 0.70430 and Bijang is
magmas. Tamura and co-workers have discussed in de- 0.70420. All these data support the conclusion that the
tail the reasons they prefer remelting of crystallized silicic magmas from Makiling had no involvement with
evolved magmas rather than melting of the lower mafic the old evolved continental crust—it can be used to
crust for the origin of the rhyolites from the Izu-Bonin support the involvement of young, mantle-derived
Arc and Daisen Volcano, Japan (Tamura and Tatsumi magmas that intruded in the lower crust. The oxygen
2002; Tamura et al. 2003). Most of their arguments isotope compositions show no evidence for hydrother-
could be applied to our present study—the inferred large mal alteration, which would support the conclusion
amounts of andesitic plutons in the underlying Luzon that shallowly intruded rocks can be ruled out (these
crust, energetic considerations of partial melting of hot shallowly intruded rocks would be subjected more to
plutons, the extreme amount of fractional crystallization hydrothermal alteration and the d18O would be shifted
needed to produce the high K2O/Na2O silicic magmas. to higher values).
It is clear that the Laguna de Bay silicic magmas are
chemically distinct from the domes and Makiling lavas,
which are nearly identical to each other (Figs. 4, 5, 6, 7). Conclusions
The silicic magmas from Laguna de Bay ignimbrites and
the domes have a source distinct from the source of the The purpose of this study is to document the relative
Makiling lavas. Adjacent to the Macolod Corridor is a abundance of silicic volcanic rocks (>65 wt % SiO2) in
sequence of Paleocene–Oligocene San Juan metavolca- the Macolod Corridor of Luzon, Philippines, and to
nics and early to mid-Miocene San Juan quartz diorite determine their origin. Geochemical data from the silicic
(Oles 1991; Sudo et al. 2001), which could potentially be domes associated with Mt. Makiling and the ash flow
a source for the silicic magmas. tuffs from Laguna de Bay Caldera reveal compositional
Tamura et al. (2003) have discussed the energy differences, demonstrating that these units cannot be
problems of silicic melts production by melting of the genetically related. However, the silicic domes and lavas
280

from Makiling Volcano have nearly identical trace Besana GM et al (1995) The shear wave velocity structure of the
element and major element trends and are considered part crust and uppermost mantle beneath Tagaytay, Philippines in-
ferred from receiver function analysis. Geophys Res Lett
of the same magmatic system. Older rhyolitic rocks 22:3143–3146
represented by core samples from 1 and 1.5 km depth Bindeman IN, Valley JW (2003) Rapid generation of both high-
beneath Makiling Volcano are also chemically similar. and low- d18O, large-volume silicic magmas at the Timber
Alteration of these rocks precludes a detailed interpreta- Mountain/Oasis Valley caldera complex, Nevada. Geol Soc Am
Bull 115:581–595
tion of their chemical affinities; nevertheless, the thickness Bindeman IN, Ponomareva VV, Bailey JC, Valley JW (2004)
of these rocks demonstrates a significant abundance of Volcanic arc of Kamchatka: a province with high- d18O magma
silicic rocks in the area. Recent experimental results sources and large-scale 18O/16O depletion of the upper crust.
(Sisson et al. 2005) have shown that the partial melting of Geochem Cosmochim Acta 68:841–865
moderate to K-rich basaltic rocks can produce high K2O/ Cardwell RK, Isacks BL, Karig DE (1980) The spatial distribution
of earthquakes, focal mechanism solutions, and subducted
Na2O silicic magmas. Our conclusion is that the silicic lithosphere in the Philippine and Northeastern Indonesian
magmas were produced by partial melting of mantle- Islands. In: Hayes DE (ed) The tectonic and geologic evolution
derived, evolved and crystallized ponded calc-alkaline of Southeast Asian seas and islands. American Geophysical
magmas. Alternatively, the silicic magmas could be Union, Washington, pp 1–35
Catane SG, Arpa MCB (1998) Large-scale eruptions of Laguna
produced by melt extraction from these partially caldera: contributions to the accretion and other geomorphic
crystallized magmas. The generation of silicic magmas in developments of Metro Manila and adjacent provinces.
island arcs in the absence of evolved continental crust Philippine Institute of Volcanology and Seismology, Diliman
represents the formation of continental crust. Catane SG, Ui T, Arpa MCB, Cabria HB, Taniguchi H (2004)
Potential hazards from the youngest explosive eruptions of
Acknowledgements We gratefully acknowledge the Philippine Laguna Caldera to metropolitan Manila, Philippines. Western
Institute of Volcanology and Seismology (PHIVOLCS) and the Pacific Geophysics Supplement 85:V33A-89
National Institute of Geological Sciences (NIGS), University of the Chiba H, Chacko T, Clayton RN, Goldsmith JR (1989) Oxygen
Philippines—Diliman, Philippines, for continued support of this isotope fractionations involving diopside, forsterite, magnetite
project. Part of this research was done while one of us (TAV) was a and calcite. Geochim Cosmochim Acta 53:2985–2995
visiting professor at NIGS. Both PHIVOLCS and NIGS provided Clayton RN, Goldsmith JR, Mayeda TK (1989) Oxygen isotope
transportation and other logistical support for the fieldwork. Jake fractionation in quartz, albite, anorthite, and calcite. Geochim
Lowenstern provided SHRIMP Pb/U zircon analyses and inter- Acta Cosmochim 53:725–733
pretation. Sam Mukasa made available unpublished Sr isotopic Cobbing EJ, Pitcher WS (1983) Andean plutonism in Peru and its
data. Unocal Philippines, Inc., provided support for the Ar/Ar and relationship to volcanism and metallogenesis at a segmented
Pb/U zircon analyses and also made available drill core samples plate edge. Mem Geol Soc Am 159:277–291
and unpublished chemical analyses. Oxygen isotopes were mea- Defant MJ, Deboer J, Oles D (1988) The Western Central Luzon
sured in John Valley’s laboratory by Jonathan Eaton. Ilya Bind- Volcanic Arc, the Philippines: two arcs divide by rifting? Tec-
eman provided insight into interpreting the oxygen isotopic data. tonophysics 45:305–317
The paper was considerably improved by thoughtful reviews by Defant M, Jacques D, Maury RC, Boer JD, Joron JL (1989)
Kent Ratajeski and Ilya Bindeman. Geochemistry and tectonic setting of the Luzon Arc,
Philippines. Geol Soc Am Bull 101:663–672
DePaolo DJ (1981) Trace element and isotopic effects of combined
wallrock assimilation and fractional crystallization. Earth
References Planet Sci Lett 53:189–202
Dimabuyu AJ, Stimac JA, Sugiaman F (2004) Detailed stratigra-
Aquino DJP (2004) Surface structure analysis for the Bulalo geo- phy of the Bulalo Geothermal Field. A tribute to Dr. Roger
thermal field: possible implications to local and regional tec- Datuin, Association of Filipinos for the Advancement of
tonics. MS, University of the Philippines, National Institute of Geoscience, p 68
Geological Sciences, Diliman, p 116 Eiler JM, Crawford A, Elliott T, Farley KA, Valley JW, Stolper
Arpa MCB, Catane SG, Mirabueno MHH (1999a) Identification EM (2000) Oxygen isotope geochemistry of oceanic arc lavas.
of large-scale eruptions of Laguna caldera from Geologic J Petrol 41:229–256
mapping of volcanic deposits found in Rizal province, Metro Fiske RS, Naka J, Iizasa K, Yuasa M, Klaus A (2001) Submarine
Manila, and adjacent areas. Philippine Institute of Volcanology silicic caldera at the front of the Izu-Bonin Arc, Japan;
and Seismology, Diliman voluminous seafloor eruptions of rhyolite pumice. Geol Soc Am
Arpa MCB, Catane SG, Mirabueno MHH, Bornas MAV (1999b) Bull 113:813–824
Identification of large-scale eruptions of Laguna caldera from Förster H, Oles D, Knittel U, Defant MJ, Torres RC (1990) The
geologic mapping of volcanic deposits found in Rizal province, Macolod Corridor—a rift crossing the Philippine Island-Arc.
Metro Manila and adjacent areas. Philippine Institute of Tectonophysics 183:265–271
Volcanology and Seismology, Diliman Gaetani GA, Kent AJR, Grove TL, Hutcheon ID, Stolper EM
Bachmann O, Bergantz GW (2003) Rejuvenation of the Fish (2003) Mineral/melt partitioning of trace elements during hy-
Canyon magma body: a window into the evolution of large- drous peridotite partial melting. Contrib Mineral Petrol
volume silicic magma systems. Geology 31:789–792 145:391–405
Bachmann O, Bergantz GW (2004) On the origin of crystal-poor Grove TL, Elkins-Tanton LT, Parman SW, Chatterjee N,
rhyolites: extracted from batholithic crystal mushes. J Petrol Muntener O, Gaetani GA (2003) Fractional crystallization and
45(8):1565–1582 mantle-melting controls on calc-alkaline differentiation trends.
Bacolcol T (2003) Etude géodésique de la faille Philippine dans les Contrib Mineral Petrol 145:515–533
Visayas. These de Doctorat, Institut de Physique du Globe de Hildreth W, Moorbath S (1988) Crustal contributions to arc
Paris, Paris, p 208 magmatism in the Andes of central Chile. Contrib Mineral
Beard JS, Lofgren GE (1991) Dehydration melting and water- Petrol 98:455–489
saturated melting of basaltic and andesitic greenstones and Hoefs J (2004) Stable isotope geochemistry. Springer, Berlin
amphibolites at 1, 3, and 6.9 kb. J Petrol 32:365–401 Heidelberg New York
 281

Knittel U, Defant MJ, Raczek I (1988) Recent enrichment in the geochemistry of ultrahigh-pressure rocks. Kluwer, Dordrecht,
source region of arc magmas from Luzon island Philippines: Sr pp 27–55
and Nd isotopic evidence. Geology 16:73–76 Pubellier M, Garcia F, Loevenbruck A, Chorowicz J (2000) Recent
Leat PT, Smellie JL, Millar IL, Larter RD (2003) Magmatism in deformation at the junction between the North Luzon block
the South Sandwich Arc. In: Leat PT (ed) Intra-oceanic and the Central Philippines from ERS-1 Images. Island Arc
subduction systems: tectonic and magmatic processes. The 9:598–610
Geologic Society, London, pp 285–313 Sisson TW, Ratajeski K, Hankins WB, Glazner AF (2005)
LeBas MJ, LeMaitre RW, Streckeisen A, Zanettin B (1986) A Voluminous granitic magmas from common basaltic sources.
chemical classification of volcanic rocks based on the total Contrib Mineral Petrol 148:635–661
alkali silica diagram. J Petrol 27:745–750 Smith IEM, Stewart RB, Price RC (2003) The petrology of a large
Le Rouzic S (1999) Seismotectonique des structures actives dans la intra-oceanic silicic eruption: the Sandy Bay Tephra, Kermadec
zone de relais, Philippines. PhD, Institut de Physique du Globe Arc, Southwest Pacific. J Volcanol Geotherm Res 124:173–194
de Paris, Paris, p 341 Sudo M, Listanco EL, Ishikawa N, Kamata H, Tatsumi Y (2001)
Listanco EL (1994) Space-time patterns in the geologic and K–Ar dating of the volcanic rocks from the Macolod Corridor
magmatic evolution of calderas: A case study at Taal Volcano, in Southwestern Luzon, Philippines: toward an understanding
Philippines. PhD Thesis, University of Tokyo, Earthquake of the Quaternary volcanism and tectonics. J Geol Soc Philipp
Research Institute, Tokyo, p 184 55:89–104
Martinez MML, Williams SN (1999) Basaltic andesite to andesite Sun S, McDonough WF (1989) Chemical and isotopic systematics
scoria pyroclastic flow deposits from Taal caldera, Philippines. of oceanic basalts: implications for mantle compositions and
J Geol Soc Philipp 54:1–18 processes. In: Norry MJ (ed) Magmatism in the Ocean Basins.
Miklius A, Flower MFJ, Huijsmans JPP, Mukasa SB, Casitillo PR Geological Society Special Publications, pp 313–345
(1991) Geochemistry of lavas from Taal volcano, southwestern Tamura Y, Tatsumi Y (2002) Remelting of an andesitic crust as a
Luzon, Philippines; evidence for multiple magma supply sys- possible origin for rhyolitic magma in oceanic arcs: an example
tems and mantle source heterogeneity. J Petrol 32:593–627 from the Izu-Bonin arc. J Petrol 43:1029–1047
Mukasa SB, Flower MFJ, Miklius A (1994a) The Nd-, Sr- and Tamura Y, Yuhara M, Ishii T, Irino N, Shukuno H (2003)
Pb-isotopic character of lavas from Taal, Laguna de Bay and Andesites and dacites from Daisen volcano, Japan: partial-
Arayat volcanoes, southwestern Luzon, Philippines: Implica- to-total remelting of an andesite magma body. J Petrol 44:2243–
tions for arc magma petrogenesis. Tectonophysics 235:205–221 2260
Mukasa SB, Flower MFJ, Miklius A (1994b) The Nd-, Sr- and Valley JW, Kitchen N, Kohn MJ, Niendorf CR, Spicuzza MJ
Pb-isotopic character of lavas from Taal, Laguna de Bay and (1995) UWG-2, a garnet standard for oxygen isotope ratios:
Arayat volcanoes, southwestern Luzon, Philippines; implica- Strategies for high precision and accuracy with laser heating.
tions for arc magma petrogenesis; geology and geophysics of Geochem Cosmochim Acta 67:3257–3266
the South China Sea and environs. Tectonophysics 235:205–221 Vielzeuf D, Holloway JR (1988) Experimental determination of the
Müntener O, Kelemen PB, Grove TL (2001) The role of H2O fluid-absent melting reactions in the pelitic system. Contrib
during crystallization of primitive arc magmas under uppermost Mineral Petrol 98:257–276
mantle conditions and genesis of igneous pyroxenites: An Villiger S, Ulmer P, Muntener O, Thompson AB (2004) The liquid
experimental study. Contrib Mineral Petrol 141:643–658 line of descent of anhydrous, mantle-derived, tholeiitic liquids
Oles D (1991) Geology of the Macolod Corridor intersecting the by fractional and equilibrium crystallization—an experimental
Bataan–Indoro island arc, The Philippines, Manila, Final study at 1.0 GPa. J Petrol 45:2369–2388
report for German Research Society Project No. Fo53/16-1 to 2 Vogel TA, Patino LC, Alvarado GE, Gans PB (2004) Silicic
and German Agency for Technical Cooperation Project No. ignimbrites within the Costa Rican volcanic front: evidence for
85.2522.2-06.100 the formation of continental crust. Earth Planet Sci Lett
Patiño Douce AE (1999) What do experiments tell us about the 226:149–159
relative contributions of crust and mantle to the origin of gra- Vogel TA et al (2006) Origin of silicic magmas along the Central
nitic magmas? In: Vigneresse JL (ed) Understanding granites: American volcanic front: Genetic relationship to mafic melts.
integrating new and classical techniques. Geological Society, J Volcanol Geotherm Res (in press)
London, pp 55–75 White AJR, Chappell BW (1983) Granitoid types and their distri-
Patiño Douce AE, McCarthy TC (1998) Melting of crustal rocks bution in the Lachlan Fold Belt, southeastern Australia. In:
during continental collision and subduction. In: Hacker BR, Roddick JA (ed) Circum-Pacific plutonic terranes. Geological
Liou JG (eds) When continents collide: geodynamics and Society of America, Boulder, pp 21–34

View publication stats