Characterization of the Angat Ophiolite: New insights from bulk major and trace element geochemistry and petrographic analysis

The Angat Ophiolitic Complex, located north-northeast of Manila, Luzon is perhaps one of Luzon’s most fundamental suspect terranes. The definite age of the ophiolite is not well-established, with various authors having different claims. Encarnacion et al. (1993) conducted a U-Pb age dating for the Angat wherein the Angat Ophiolite was found to have an Early Middle Eocene age (48 Ma), close to the ZOC age - leading to the assumption that the two ophiolites are related and underlie most of Central Luzon. Hence, this paper aims to collect samples from the Eocene and Cretaceous Angat Ophiolite for petrographic and geochemical analyses, identify the relationship between the separated basalt units from the basalts in the main ophiolite body, and compare the geochemistry of the Eocene and Cretaceous units of the Angat Ophiolite with the other well-known ophiolites from the literature. Field investigations were conducted in five important localities wherein 15 samples were obtained. The results highlighted that the major mineral phases in the samples collected are plagioclase and augite, with some samples having hornblende and olivine. All samples appear to have undergone low-grade greenschist facies metamorphism, which may be attributed to hydrothermal alteration. Chloritization of pyroxene minerals is also evident, along with the hydrothermal alteration products of plagioclase and pyroxenes. The relationship between the EAO, MOC, and separated basalt patch in Marilaque Highway is identified. The Coto Block of the Zambales Ophiolite and the Angat Ophiolite may be related and formed over a mantle initially enriched by a subduction component.


Introduction
The Angat Ophiolite located north-northeast of Manila, Luzon, forms part of the southern Sierra Madre range.It is a north-south trending, east-dipping ophiolitic complex consisting of dismembered bodies of pillow basalts, basalt breccias, flows, sediments, diabase dikes, and gabbro [1,2].Available major and trace element geochemistry reveals that the Angat Ophiolite rocks exhibit Mid-Ocean Ridge Basalt -Island Arc Tholeiite (MORB-IAT) characteristics, which points to a subduction-related origin for the ophiolite [1,3].However, a supra-subduction zone being the environment of formation is not definitively indicated by the MORB-IAT characteristics of the Angat Ophiolite.A Late Cretaceous age was given for the Angat Ophiolite based on its proximity to Late Cretaceous strata, along with the sedimentary rocks associated with its pillow basalts (e.g., [4,5]).However, mapping the Montalban area revealed that the ophiolite is cut by several faults that complicate paleontological and sedimentologicalbased dating [1].By 1993, Encarnacion et al. conducted a U-Pb age dating for the Angat; the Angat Ophiolite was found to have an Early Middle Eocene age (48 Ma), close to the Zambales Ophiolitic Complex (ZOC) age [6].This led Encarnacion to assume that the two ophiolites are related and underlie most of Central Luzon.An issue with this, however, is the lack of geochemical data for the Angat and the difference in sedimentary cover over the two ophiolites.Accordingly, Angat's age restriction remained a problem due to the lack of radiometric age information for the pillow basalt portion of the Angat.Furthermore, there is still no discussion of the connection between the Angat and isolated basaltic rocks, especially those exposed along the Marilaque Highway.
Hence, this study aims to identify a geochemical relationship between the Eocene and Cretaceous units of the Angat Ophiolite to understand further its complexity, which will then be compared with other well-known ophiolites through literature.In particular, the objective of the study are (1) to collect samples from the Eocene and Cretaceous Angat Ophiolite for petrographic and geochemical analyses; (2) identify the relationship between the separated basalt units from the basalts in the main ophiolite body; and (3) compare the geochemistry of the Eocene and Cretaceous units of the Angat Ophiolite with the other well-known ophiolites from the literature, particularly the Zambales Ophiolite.

Methodology
This study aims to characterize the Angat Ophiolite, which forms the southern portion of the Sierra Madre range, through the geochemical and petrographic analysis of its gabbro, basalt, and diabase units (Figure 1).Fieldwork observations will provide the data for the lithology, locality, and degree of weathering.The samples collected will then be analyzed for bulk major and trace element geochemistry, mineral content and textural relationships.Geochemical analyses were done through submitting the samples in Intertek Philippines.Data acquired were plotted through GCDkit program [7].Meanwhile, petrographic analysis was performed using a petrographic microscope available at Mapua University.The variations in the rock types from the different sample sites were revealed through transmitted light microscopy.Focused petrographic examinations were performed on the thin section slides to identify the primary and secondary components and their textures.

Field Investigation
Field investigations were conducted in five important localities across the Angat Ophiolite.All data used in the study were grouped according to their significant localities.The samples were divided into groups based on their locality relative to the East Marikina Fault and are shown in Figure 2: (1) The Eocene Angat Ophiolite (EAO) refers to the samples collected west of the fault.These were collected along the Lacotan and Hanginan Rivers (LR and HR samples, respectively) and represent the gabbro and sheeted dike section.
(2) The Montalban Ophiolitic Complex (MOC) refers to the samples collected east of the fault.Samples were acquired from the Mango and Nangka (COGEO) rivers (MR and NR samples, respectively).These sections represent the supposed pillow basalt section of the Angat.
(3) The Marilaque Highway (MH) sample represents the basalt sample obtained in one of the patches separated from the main body of the ophiolite, located in Pinugay, Rizal, along the Marilaque Highway.

Petrographic Interpretation
The major mineral phases in the samples collected from the mafic constituents of the Angat Ophiolitic Complex are plagioclase and augite.Hornblende and olivine are noted in some samples.All samples appear to have undergone low-grade greenschist facies metamorphism, which may be attributed to hydrothermal alteration.Chloritization (partial and complete) of pyroxene minerals (augite) is also evident, along with the hydrothermal alteration products of plagioclase and pyroxenes (through saussuritization).It is important to note that the alteration to chlorite may only be due to postdepositional hydrothermal solutions and weathering.The typical assemblages in the samples include chlorite and carbonates.
Plagioclase crystals typically show little to no alteration, with some samples exhibiting alteration to illite fibers, illite ± epidote, or calcite.Chlorite, illite, calcite, and opaques are the common secondary minerals in all the collected samples.Limonite, chalcedony, epidote, clinozoisite, and zeolite are also noted under secondary minerals.Textures of the basalt samples vary from aphyric and porphyritic upon observation.Meanwhile, the gabbro samples exhibit hypidiomorphic, xenomorphic granular, and hypidiomorphic granular textures.

Whole-rock Major Element Geochemistry
Generally, the tholeiitic trend of the data from both ophiolites was supported by the fractionation trend illustrated in Figure 3.The positive correlation between FeOt/MgO and TiO2 is commonly observed in tholeiitic magma [8].The Harker-type diagram also supported the crystallization order of the Angat and ZOC magma (Figure 3).The negative correlation in CaO, Al2O3, and MgO vs. FeOt/MgO may suggest that olivine was the first mineral to form, followed by plagioclase.The hooked trend in the FeOt vs. FeOt/MgO may suggest that clinopyroxene formed after plagioclase.This is consistent with the (Ol)-Pl-Cpx crystallization order observed in the Angat petrography.This trend is also present in the Coto and is consistent with the (Ol)-Pl-Cpx crystallization order noted by previous studies [3,9,10].One can observe that the trend of the Acoje in the FeOt vs. FeOt/MgO dipped earlier than the Angat and Coto.Also, the hooked trend of the Acoje in the Al2O3 vs. FeOt/MgO plot may suggest that the first mineral to form after the olivine was pyroxene, followed by plagioclase.This is consistent with the (Ol)-Px-Pl crystallization order observed by [9,10].Rock type discrimination diagrams that used immobile trace elements as proxy for the conventional Na2O+K2O were utilized considering the altered state of the samples.Figures 4A and 4B revealed that most of the samples from the EAO plot in the sub-alkaline basalt field, while the MOC samples are scattered, with some plotting in the sub-alkaline basalt field, while the others plot in the andesite and trachyandesite fields [11].The scattering of data within certain plots, such as in SiO2, Al2O3, and Na2O may have been caused by element remobilization during alteration (Figure 5) [12].

Tectonic Discrimination for the Angat and ZOC
The use of immobile trace elements in tectonic discrimination diagrams has been a common practice regarding ophiolite geochemistry (e.g., [2,3,13,14]) due to these elements being immobile during seafloor alteration or low-grade metamorphism [15].The MORB-IAT characteristic of the Angat could be seen in the Ti vs. V diagram (Figure 6), where the EAO, MOC, and MH samples plotted in both MORB and IAT fields (Ti/V = 20 -50 and Ti/V > 10, respectively) [16].The Angat data and literature coincided with the Coto rocks, while the Acoje plot mostly in the Ti/V = 10 field.Aside from this, the Zr-Th-Nb diagram revealed a distinction between the datasets (Figure 7) [17].It showed that the EAO and MOC samples plotted along the MORB field, while the Acoje Boninites plotted mostly in the IAT field.The Angat literature was plotted in the Enriched Mid-Ocean Ridge Basalt (E-MORB)/Within-Plate Basalt (WPB) field, while one MH literature sample was plotted near the calc-alkaline basalt (CAB) field.This may be due to the limited subduction components during the formation of the Angat.The detail of this shall be discussed in another section.

Relationship between the EAO, MOC, and MH.
It is established that the EAO and MOC are part of one ophiolite suite in the form of the Angat Ophiolite.This is evident in the consistent plotting of the present study's samples.Trace element ratio diagrams (Figures 8 and 9) further delineate this relationship.Accordingly, a fractionation trend between the EAO and MOC can be observed in the Harker-type diagrams, which may have been produced by fractional crystallization upon magma rising and reflects the crystallization order observed in the petrography.
Considering that the two units were cut by a strike-slip fault (i.e., the East Marikina Valley Fault), the two sections could still be correlated based on the presence of sheeted dikes on both sides of the fault.The relation between the two units would also impart a minimum Early Middle Eocene age for the Lower Pillow Basalts in the MOC.However, the "Upper" pillow basalts [18] may be older but are limited to at least the Middle to Late Paleocene.
Determining the relationship of the MOC and MH was challenging due to the sporadic plotting of the MH samples and literature.MH literature data constantly plotted away from the trend of the Angat, which plotted more closely to the arc-like range of the Acoje.This difference was further demonstrated by the contrasting trend of the MH in the La/Sm (N) vs. Ba/La diagram (Figure 8B).Hence, the MH basalt may be genetically and structurally unrelated to the Angat Ophiolite due to its spatial and geochemical difference from the MOC.

Relationship between the Angat and ZOC
The consistent plotting of the present study's samples in Figure 6, along with the similar crystallization order of (Ol)-Pl-Cpx [3,10], trace element ratio diagrams, and fractionating trends between the Angat and Coto, suggest that the two units may be related and formed from a similar magmatic process.Geary et al. originally proposed that the Coto experienced two stages of crustal formation: an initial N-MORB dominant magmatism in possibly a large oceanic or back-arc spreading center followed by the crustal modification of incipient island arc magmatism, possibly in a proto-forearc setting [19].IAT dike intrusions in the mantle section of the Coto and the progressive depletion of the mantle wedge from the Coto to the Acoje may have been manifestations of this process [3,20].
Based on Figure 9, the Angat-Coto may have been associated with the magmatism of the Sierra Madre, given that the trace element signatures of the Angat, Coto, and Caraballo Formation overlap and share similar trends.It is possible that the subducted slab composition was preserved in the Eocene volcaniclastics (i.e., Caraballo and its Eocene equivalents).Previous studies found that these units were the result of the magmatism in Luzon during the Early Cenozoic (e.g., [21,22]).Also, the sharp contrast between the data trends suggests that a different subducted slab must have influenced the Acoje.This relationship could be observed in the Acoje, Angat, Caraballo, and Coto's trend, sharply contrasting the back-arc Lau Basin trend.

Conclusions
(1) Given the various degrees of alteration in the nature of the rocks, significant chlorite-epidote alteration in the phenocrysts and groundmass of the samples is evident.This suggests chemical alteration through iron and magnesium-bearing hydrothermal fluids.
(2) A homogeneous petrographic character is exhibited in the EAO and MOC samples due to the plagioclase-clinopyroxene phenocryst mineralogy of evolved MORB magmas [23] (Grove and Bryan, 1983).Accordingly, the mineralogic and textural relationships of the samples can be related to cooling history and provide information on magma chemistry.
(3) The previous conclusion is further supported by the fractionation trend observed in the Harkertype bivariate plots, showing an apparent decrease in MgO with increasing TiO2, typical of tholeiitic magmas; (4) The study confirms that EAO and MOC are part of one ophiolite suite in the form of the Angat Ophiolite.This is evident in the consistent plotting of the present study's samples in numerous tectonic discrimination diagrams and trace element ratio bivariate plots.
(5) The Coto Block of the Zambales Ophiolite and the Angat Ophiolite may be related and formed over a mantle initially enriched by a subduction component.This conclusion is supported by the similar crystallization order of (Ol)-Pl-Cpx [3,10], consistent overlapping of plots and trends, and similar enrichment pattern range.

Recommendations
(1) It is advised to increase the number of samples and conduct a test on the complete set of trace elements, particularly REEs, for the Angat Ophiolite since it still lacks the complete trace element data to establish its chemical affinity.
(2) Sampling should be accomplished in areas far from the East Marikina Fault since the exposures where the samples were collected in this study were heavily affected by shearing and/or alteration.
(3) The mapping of exposures of the Angat Ophiolite should be updated, along with further mapping of the northern part of Montalban.
(4) Updated isotope dating should be conducted to verify the ages given for the Angat Ophiolite.

Figure 3 .
Figure 3. Harker-type bivariate diagram for Acoje, Coto, EAO, MOC, MH, and Angat Literature data.Red lines denotes possible fractionation trends for Angat while purple lines denote Acoje trend.Legends are similar with Figure 4.

Figure 8 .
Figure 8. Trace element ratio bivariate diagrams.(A) and (B) were modified from Castillo and Newhall (2004), while (C) was from Li et al. (2002).Violet arrow for Acoje trend; the black arrow for Coto trend; Yellow arrow for MH trend; green arrow for Angat trend.La/Sm (N) are chondrite normalized after Sun and McDonough (1989).Legends same with Figure 9.

Figure 9 .
Figure 9. Trace element bivariate plot.Arrows symbol similar with Figure 9; cyan arrow for Caraballo trend.The orange field are samples compiled from the ODP Leg 135 on the Lau Basin.