Portal de Periódicos da CAPES

Zircon growth in (U)HP quartzo-feldspathic host gneisses exhumed in the Woodlark Rift of Papua New Guinea

2013; Wiley; Volume: 15; Issue: 4 Linguagem: Inglês

10.1002/2013gc004964

1525-2027

N. Alex Zirakparvar, Suzanne L. Baldwin, Axel K. Schmitt,

High-pressure geophysics and materials

Geochemistry, Geophysics, GeosystemsVolume 15, Issue 4 p. 1258-1282 Research ArticleFree Access Zircon growth in (U)HP quartzo-feldspathic host gneisses exhumed in the Woodlark Rift of Papua New Guinea N. A. Zirakparvar, Corresponding Author N. A. Zirakparvar Department of Earth Sciences, Syracuse University, Syracuse, New York, USA Present address: Department of Earth and Planetary Sciences, American Museum of Natural History, Central Park West 79 St., New York, New York, 10024 USACorrespondence to: N. A. Zirakparvar, [email protected]Search for more papers by this authorS. L. Baldwin, S. L. Baldwin Department of Earth Sciences, Syracuse University, Syracuse, New York, USASearch for more papers by this authorA. K. Schmitt, A. K. Schmitt Department of Earth and Space Sciences, University of California, Los Angeles, California, USASearch for more papers by this author N. A. Zirakparvar, Corresponding Author N. A. Zirakparvar Department of Earth Sciences, Syracuse University, Syracuse, New York, USA Present address: Department of Earth and Planetary Sciences, American Museum of Natural History, Central Park West 79 St., New York, New York, 10024 USACorrespondence to: N. A. Zirakparvar, [email protected]Search for more papers by this authorS. L. Baldwin, S. L. Baldwin Department of Earth Sciences, Syracuse University, Syracuse, New York, USASearch for more papers by this authorA. K. Schmitt, A. K. Schmitt Department of Earth and Space Sciences, University of California, Los Angeles, California, USASearch for more papers by this author First published: 26 November 2013 https://doi.org/10.1002/2013GC004964Citations: 9AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Abstract To understand zircon behavior as a function of bulk composition and metamorphic grade in the world's youngest (U)HP terrane, we report U-Pb SIMS spot-mode and depth-profiling analyses for quartzo-feldspathic gneisses. Zircons from two gneisses contain Cretaceous inherited cores, with μm sized metamorphic rims requiring depth profiling for reliable dating. Linear regression of the common-Pb uncorrected data for rims yield 206Pb/238U ages of 2.89 ± 0.29 Ma and 2.77 ± 0.99 Ma (concordia intercept ages at 95% confidence). The older age is within two-sigma error of previously reported 206Pb/238U ages on zircons from mafic eclogite within the gneiss, indicating that rims formed on inherited cores within host gneisses during eclogite facies metamorphism. At the (U)HP locality zircons from host gneiss lack inheritance and yield a 206Pb/238U age of 3.66 ± 0.13 Ma. These results are younger than previously reported 206Pb/238U ages on zircons from coesite eclogite within the gneiss, but are within error of the youngest reported LA-ICP-MS 206Pb/238U zircon ages on retrogressed mafic eclogite. We also report intragrain geochemical heterogeneity, indicated by zircon Hf, Y, and Ti variations in depth profiles which suggest chemical disequilibrium over the interval of zircon growth. Collectively, these results indicate that zircon recrystallization and new growth of zircon rims on relict grains occurred during eclogite facies metamorphism and during subsequent retrogression, but not at (U)HP conditions. Comparison between results from PNG and other (U)HP terranes bolsters previous suggestions that the PNG (U)HP terrane evolved rapidly. Key Points Zircon grows in host gneiss over a wide range of conditions SIMS depth profiling reveals um scale geochemical heterogeneity The rates of processes in the PNG (U)HP terrane are similar to the Western Alps 1. Introduction Ultrahigh-pressure ((U)HP) metamorphic terranes comprise complex assemblages of highly deformed mafic and quartzo-feldspathic rocks. The P-T-t evolution of quartzo-feldspathic host gneiss relative to enclosed mafic eclogite is often poorly understood because diagnostic major (e.g., omphacite and almandine-pyrope garnet) or trace (e.g., coesite or microdiamond) (U)HP mineral assemblages are often absent, or not preserved in quartzo-feldspathic host gneisses. In both active and ancient tectonic settings, understanding the evolutionary paths of mafic eclogites and felsic host gneisses is a crucial piece of information in modeling the (U)HP terrane's rheologic properties at depth [e.g., Brownlee et al., 2011], which in turn has implications for interpreting geophysical data and developing geodynamic models. Whether host gneisses were metamorphosed simultaneously with mafic eclogite can be addressed by combining geochronologic and trace element data [e.g., Liati, 2005]. For this purpose, zircon has been used extensively because it can form entirely new crystals or grow as rims on preexisting grains during metamorphism [Rubatto et al., 1998; Liati, 2005; McClelland et al., 2009; Gao et al., 2011; McClelland and Lapen, 2013], and trace element systematics in zircons can be interpreted with respect to mineral reactions that change the bulk distribution coefficients during zircon growth [Rubatto, 2002; Hanchar and VanWestrenen, 2007; Harley et al., 2007; Harley and Kelly, 2007; Rubatto and Hermann, 2007; Scherer et al., 2007; Monteleone et al., 2007]. However, the fact that zircon forms under a wide range of P-T conditions in silica-rich rocks also lends complexity to the interpretation of U-Pb zircon ages during mineral paragenesis in (U)HP quartzo-feldspathic gneisses. The purpose of this study is to examine the U-Pb age and trace element distributions in zircons from three quartzo-feldspathic gneisses (Figure 1) in the (U)HP terrane exhumed in the western Woodlark Rift of Papua New Guinea (PNG). In this rapidly evolving plate boundary zone, extensional gneiss domes have exposed the world's youngest known (U)HP rocks and research has focused on understanding the geologic history of this region at scales ranging from the entire Australia-Pacific (AUS-PAC) plate boundary down to the multistage evolution of individual minerals [Baldwin et al., 1993, 2004, 2008; Taylor et al., 1995; Monteleone et al., 2001, 2007; Little et al., 2011; Zirakparvar et al., 2011, 2012; Gordon et al., 2012]. This pre-existing body of research provides a framework within which to interpret the U-Pb age and trace element results from zircons in quartzo-feldspathic host gneisses, thereby allowing us to relate the observed U-Pb age and trace element data to specific processes in the tectonic evolution of the PNG (U)HP terrane. Figure 1Open in figure viewerPowerPoint Geologic map of D'Entrecasteaux Islands region with sample localities and showing major lithological [Monteleone et al., 2007; Baldwin et al., 2008; Davies, 2012], structural [Little et al., 2007], and topographic features [Miller et al., 2012] of the region. Abbreviations are as follows: PNG = Papua New Guinea; G.I. = Goodenough Island; F.I. = Fergusson Island; N.I. = Normanby Island. Zircons in two samples from the Woodlark Rift exhibit internal morphologies, as revealed by cathode-Ray Luminescence (CL) imaging of polished zircon cross sections, consistent with an interpretation of older cores overgrown by younger rims (Figure 2). The rims are typically <5 µm thick, necessitating a sampling and analysis strategy with sufficient spatial resolution to fully resolve the U-Pb age and trace element systematics at the micron scale. In this study, Secondary Ionization Mass Spectrometry (SIMS) was used in spot mode to acquire U-Pb age data from polished zircon cross sections (Figure 3a). In addition, unpolished external crystal surfaces (Figure 3b) were analyzed via depth profiling [Breeding et al., 2004; Trail et al., 2007; Gordon et al., 2009a, 2009b; Zou et al., 2010; Schneider et al., 2011] at variable depths (up to 15 µm) to acquire U-Pb age and Ti, Hf, and Y concentration data at the submicron-scale resolution afforded by this technique (Figure 3c). Figure 2Open in figure viewerPowerPoint Selected cathode-ray luminescence (CL) images for zircons extracted from gneisses in the D'Entrecasteaux Islands. Note the up to 5 µm thick dark CL overgrowths for zircons from sample PNG 03-118m. Zircons from sample PNG 06-21a also exhibit dark CL overgrowths, although they are much thinner than in sample 03-118m and not visible in this image. Zircons from sample PNG 08-10g do not exhibit any dark CL overgrowths. See text for discussion of ages for zircons from these samples. Figure 3Open in figure viewerPowerPoint Schematic diagram illustrating different types of SIMS analyses ((1) spot mode for polished cross sections, (2) spot mode for unpolished external surface, and (3) depth profiling) performed in this study. A comparison of the SIMS data acquired in this study with preexisting geochronology allows for an examination of the history of the quartzo-feldspathic gneisses relative to the prograde, peak (U)HP, and retrograde metamorphism as recorded by zircon U-Pb systematics. The SIMS depth-profiling results also provide unique insights into the spatial scale of trace element heterogeneity during zircon crystallization in a former subduction complex that is now being rifted apart. Lastly, the data acquired in this study facilitate detailed comparisons between the time scales of zircon crystallization in the PNG (U)HP metamorphic terrane and those reported from other (U)HP terranes globally [e.g., McClelland and Lapen, 2013]. 2. Geologic Background and Samples Analyzed Geologic Overview Three quartzo-feldspathic gneiss samples were chosen for detailed SIMS analysis in this study. Samples were collected from extensional gneiss domes [Hill, 1994; Little et al., 2007, 2011] in the D'Entrecasteaux Islands, southeastern PNG (Figure 1), which occur in a zone of active extension (i.e., Woodlark Rift) at the western apex of the Woodlark Basin. Samples are described in detail in section 2.2, whereas this section provides a general geologic overview of the region. Westward propagating seafloor spreading in the Woodlark Basin has occurred since ∼6 Ma [Taylor et al., 1995], and the active seafloor-spreading rift tip is currently situated offshore northeast Normanby Island (Figure 1). In the D'Entrecasteaux Islands, Cretaceous aged volcaniclastic sediments and basalts derived from the Gondwana rifted margin [Zirakparvar et al., 2012] were subducted, metamorphosed at (U)HP conditions in the Late Miocene to Pliocene [Monteleone et al., 2007; Baldwin et al., 2008; Zirakparvar et al., 2011], and since then have been exhumed to the surface during rifting [Little et al., 2007, 2011; Martinez et al., 2001; Webb et al., 2008]. Remnants of subducted lithosphere are found in the core zones and shear zone carapaces of the D'Entrcasteaux Islands domes, where felsic and intermediate gneisses encapsulate mafic eclogites with Late Miocene–Pliocene metamorphic crystallization ages. U-Pb zircon and Lu-Hf garnet dating of the mafic eclogites [Baldwin et al., 2004; Monteleone et al., 2007; Zirakparvar et al., 2011; Gordon et al., 2012], and P-T constraints [Davies and Warren, 1992; Hill and Baldwin, 1993; Baldwin et al., 2008], indicate that coesite eclogite in the D'Entrecasteaux Islands was exhumed from depths of at least 90 km since ∼8 Ma. K/Ar, 40Ar/39Ar, and fission-track dating techniques applied to the lower-plate rocks have also documented extremely rapid (e.g., ≥100°C/Myr) cooling [Baldwin et al., 1993]. It is not yet known when these rocks were subducted, but it is noteworthy that a large garnet porphyroblast in the shear zone carapace on Goodenough Island yielded a 68 ± 3.6 Ma Lu-Hf age [Zirakparvar et al., 2011] coinciding with the timing of ophiolite obduction on the Papuan peninsula [Lus et al., 2004] and the age of diabase drilled on Ocean Drilling Program (ODP) Leg 180 from the Moresby Seamount [Monteleone et al., 2001]. In the D'Entrecasteaux Islands, seismic data [Abers et al., 2002], stream profile analysis [Miller et al., 2012], and thermochronologic data [Baldwin et al., 1993], all suggest that exhumation of lower-plate rocks occurred during Plio-Pleistocene to Holocene time and may still be ongoing. Samples Analyzed Two of the quartzo-feldspathic gneiss samples (PNG 03-118m and PNG 08-10g; Figure 1) are core zone rocks hosting Miocene–Pliocene mafic eclogites, whereas a third sample (PNG 06-21a; Figure 1) is from the Wakonai extensional shear zone carapace separating core zone rocks from the unmetamorphosed upper plate on northern Goodenough Island [Baldwin et al., 1993; Hill, 1994; Little et al., 2011]. Published data provide a framework within which to interpret the SIMS zircon U-Pb and trace element data acquired in this study. These preexisting constraints are summarized here and in Table 1. Table 1. Coordinates and published P-T-t constraints for samples analyzed in this studya PNG 03-118m S 9° 29.167′ For mafic eclogite boudin within sample PNG 03-118m E 150° 14.750′ Felsic host gneiss: Qtz + Pl + Ms + Am bZircon 206Pb/238U age via SIMS = 2.09 ± 0.49 Ma interpreted as crystallization at eclogite or amphibolite conditions. Zr in Rt therm = 677 to 817°C; Jd barom = min P of 14 kbar. PNG 08-10g S 9° 29.319′ For coesite eclogite boudin within sample PNG 08-010g E 150° 27.696′ Felsic host gneiss at coesite locality: Qtz + Pl + Ms +Bt bZircon 206Pb/238 U age via SIMS = 7.9 ± 1.9 Ma and cLu-Hf garnet age = 7.1 ± 0.7 Ma; both interpreted as recording crystallization at (U)HP conditions. bZr in Rt therm. = 650–680°C; Jd barom. = min P of 15 kbar. dAmphibole from retrogressed rind 40Ar/39Ar age = 9.83 Ma. Muscovite from dike at same outcrop 40Ar/39Ar age = 3.52 ± 0.10 Ma. U-Pb results for mafic eclogite boudin without documented coesite eZircon 206Pb/238 U ages via LA-ICP-MS = 9.1 ± 0.6 Ma to 3.8 ± 1.0 Ma in retrogressed rim of eclogite, 7.4 ± 1.1 Ma to 4.1 ± 1.3 Ma in unretrogressed interior. Zircon 206Pb/238U ages via CA-TIMS = 5.82 ± 0.02 Ma to 4.78 ± 0.17 Ma. PNG 06-21a S 9° 18.831′ For rocks in same shear zone as sample PNG 06-21a E 150° 17.430′ Shear zone gneiss: Qtz + Pl + Ms + Grt + Chl 40Ar/39Ar ages for minerals in nearby samples from Wakonai shear zone: dmuscovite = 1.5 Ma, biotite = 1.4 Ma, K-feldspar = 1.4 Ma. Apatite fission track for rocks in Wakonai shear zone = 0.8 Ma. cSample <5 km away in Wakonai shear zone records garnet growth at 68 ± 3.6 Ma. a The precise location and mineralogy of the three samples examined in this study is also provided. Mineral, and other, abbreviations are as follows: Qtz = quartz, Pl = plagioclase feldspar, Ms = muscovite, Am = amphibole, Bt = biotite, Grt = garnet, Zrn = zircon, Chl = chlorite, Kfs = K-feldspar, Zr = zirconium, Rt = rutile, Jd = jadeite. Sources of data are as follows: b Monteleone et al. [2007]; c Baldwin et al. [1993]; d Zirakparvar et al. [2011a]; e Gordon et al. [2012]. 2.2.1. Gneissic Host of ∼2 Ma Eclogite: PNG 03-118m Sample 03-118m is a quartzo-feldspathic host gneiss from southeastern Goodenough Island (Figure 1). At outcrop scale, this gneiss encapsulates mafic eclogite that occurs as a meter-scale lenticular boudin. The eclogite preserves the peak metamorphic assemblage garnet + omphacite + phengite + rutile whereas gneiss sample 03-118m contains an assemblage of quartz + plagioclase + muscovite + biotite + K-feldspar. Zircon crystals from one of these mafic eclogite boudins (sample 03118b in Monteleone et al. [2007]), occur as inclusions in garnet and coexist with omphacite in the rock matrix. These zircons yielded a SIMS concordia intercept 206Pb/238U age of 2.09 ± 0.49 Ma (through regression of common-Pb uncorrected data with unpinned upper intercept; mean square of weighted deviates [MSWD] = 3.3) [Monteleone et al., 2007]. Trace element data and the coexistence of the dated zircons with omphacite as inclusions in garnet, was used to interpret this U-Pb age as the time of eclogite facies metamorphism, with the authors noting that zircon growth under amphibolite facies conditions cannot be ruled out. Zirconium-in-rutile thermometry for this mafic eclogite gave temperatures ranging from 677 to 817°C, and jadeite barometry gave minimum pressures of 14 kbar [Monteleone et al., 2007]. Zircons extracted from the host gneiss examined in this study (PNG 03-118m) range in length from 100 to 200 µm, are prismatic to acicular, and exhibit oscillatory zoned metamict cores mantled by dark CL rims of variable thickness and up to 5 µm in wide (e.g., Figure 2). 2.2.2. Host Gneiss From the Coesite Eclogite Locality: PNG 08-10g Sample 08-10g is a quartzo-feldspathic gneiss containing quartz + plagioclase + muscovite + biotite from the coesite eclogite locality (Figure 1) [Baldwin et al., 2008]. The coesite eclogite preserves the peak (U)HP mineral assemblage (clinopyroxene+ garnet+ phengite+ coesite + quartz + hornblende + zoisite + apatite + rutile + zircon; sample 89321 in Monteleone et al. [2007], Baldwin et al. [2008], and Zirakparvar et al. [2011]). A 206Pb/238U age of 7.9 ± 1.9 Ma (regression with unpinned upper intercept) based on in situ SIMS U-Pb analysis of zircon in garnet [Monteleone et al., 2007], and a Lu-Hf garnet age of 7.1 ± 0.7 Ma [Zirakparvar et al., 2011] have both been interpreted to record crystallization of this coesite eclogite at (U)HP conditions. The amphibolite rind of the coesite-bearing eclogite yielded an older 40Ar/39Ar amphibole inverse isochron age of 9.83 Ma (sample 89321 in Baldwin et al. [1993]). The presence of phengite and amphibole, as well as a muscovite 40Ar/39Ar age of 3.52 ± 0.10 Ma, from the same locality (sample 89320 in Baldwin et al. [1993]), indicates that hydrous phases were present throughout the evolution of this (U)HP terrane. Mafic eclogite (sample PNG 08-10f in Brownlee et al. [2011] and Gordon et al. [2012]), occurs as a boudin within host gneiss sample PNG 08-10g. Although coesite has yet to be documented within this mafic eclogite sample, recent work by Gordon et al. [2012] reports a range of in situ laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) 206Pb/238U zircon ages of 9.1 ± 0.6 to 3.8 ± 1.0 Ma in the retrogressed part of this eclogite and 7.4 ± 1.1 to 4.1 ± 1.3 Ma in the "fresh" part of the eclogite, with no inheritance reported. Gordon et al. [2012] also report chemical abrasion-thermal ionization mass spectrometry (CA-TIMS) 206Pb/238U ages between 5.82 ± 0.2 and 4.78 ± 0.17 Ma for zircons separated from the mafic eclogite. This data suggests that zircon crystallization in mafic eclogite at the coesite locality occurred over a range of P-T conditions from 9.1 to 3.8 Ma. However, the depths corresponding to zircon crystallization during exhumation are not well known. Zircons in felsic host gneiss PNG 08-10g are prismatic and acicular, up to ∼400 µm in length and ∼170 µm in width. All of the zircons in this sample exhibit complex internal zoning (Figure 2). Some of the grains have convoluted metamict cores surrounded by concentrically zoned regions, and others exhibit a completely chaotic internal morphology lacking any discernable patterning. The zircons from this sample lack the conspicuous dark CL overgrowths observed in PNG 03-118 m. 2.2.3. Gneiss in Dome Bounding Shear Zone: PNG 06-21a Sample PNG 06-21a is from an outcrop of layered quartzo-feldspathic gneiss in the Wakonai shear zone (Figure 1), northern Goodenough Island [e.g., Baldwin et al., 1993; Hill, 1994; Little et al., 2011]. Sample PNG 06-21a contains quartz + plagioclase + muscovite + zircon + garnet + chlorite. Integrated 40Ar/39Ar ages for a sample in close proximity to 06021a are ∼1.5 Ma for white mica, ∼1.4 Ma for biotite, and ∼1.4 Ma for K-feldspar [Baldwin et al., 1993]. Apatite fission-track dating for rocks in the Wakonai shear zone yielded ages of ∼0.8 Ma [Baldwin et al., 1993]. These young 40Ar/39Ar and apatite fission track ages indicate that this extensional shear zone has been active throughout the last few Myr and is intrinsically related to the exhumation of (U)HP metamorphic rocks in the western Woodlark Basin. Sample PNG 06-21a contains both prismatic zircons with lengths of ∼60 to ∼200 µm and subrounded zircons with diameters of ∼30 to ∼100 µm. Most of the grains exhibit simple to complex oscillatory magmatic zoning. A thin (e.g., <2 µm) light gray CL overgrowth occurs on some grains from this sample (Figure 2). 3. Secondary Ionization Mass Spectrometry (SIMS) Analytical Methods Ion microprobe measurements were conducted at the UCLA SIMS lab using the CAMECA ims 1270 high-resolution, high-sensitivity ion microprobe. Resolution of mass interferences within the mass range analyzed was possible due to the instrument's high mass resolution (∼4500 m/Δm). The data reported in this paper was collected during three sessions (August 2008, July 2010, and May 2011). The first session consisted of analyses of polished zircon cross sections performed in spot mode (Figure 3a). The second analytical session targeted unpolished zircon grain surfaces (Figure 3b). For the third analytical session, analyses also started at the exterior surface of unpolished zircon grains, but profiles had a longer duration and penetrated up to ∼15 µm into the crystal (Figure 3c). Only U-Pb analysis was performed during sessions one and two, whereas Ti, Hf, and Y concentrations were measured in conjunction with U-Pb isotopic data during the third session. Following extraction from ∼2 of host rock using standard protocols, including methylene iodide gravimetric separation, zircons were prepared for SIMS analyses in two ways. For analysis of the polished cross sections, grains were mounted in epoxy along with the zircon standard AS3 [Paces and Miller, 1993]. These mounts were polished, exposing the internal surfaces of the grains, prior to the application of a thin coating of carbon for scanning electron microscope (SEM) CL imaging. Prior to SIMS analysis the carbon coating was removed through light polishing, the mount was cleaned with dilute HCl, and then coated with an ∼30 nm Au film. For analyses on unpolished grain surfaces zircons were pressed into indium metal along with AS3 zircon standard (1099.1 Ma) [Paces and Miller, 1993]. Mounts were cleaned with dilute HCl and also coated with an ∼30 nm Au film. 57Fe and 197Au were analyzed during both types of SIMS analysis to monitor beam overlap onto Fe-rich and Ti-rich inclusions in zircon, and surface-derived contaminants, respectively. The analytical parameters for two types of SIMS analyses and data processing are described below. Spot Mode (Sectioned Crystals and Unpolished Rims) A 12.5 kV primary 16O− beam with a ∼20 nA current and ∼25 µm beam diameter was used for zircon excavation. Intensities of monatomic U+, Th+, and Pb+ ions and 94Zr2O+ and UO+ molecular ions were measured with a discrete dynode electron multiplier in peak jumping mode. Individual analyses consisted of 15 cycles. O2 flooding at pressures of 3 × 10−5 Torr was applied to the sample surface to enhance Pb yield. In-house software (ZIPS v. 3.0.4; developed by Chris Coath), was used to correct intensity ratios for drift and relative sensitivity variations using a linear fit in UO+/U+ versus Pb+/U+. Depth Profiling Primary ion beam conditions for long depth profiles (session three) were similar to those described in section 3.1, but depth resolution was enhanced by narrowing the field aperture in the secondary ion path to block ions from the periphery of the analysis crater. In addition to the ion species used in spot mode, the petrogenetically relevant trace elements Ti+, Y+, and HfO+ were added to the analysis mass table. Ti and Hf abundances in zircon reflect crystallization temperature (Hf indirectly because of its enrichment during zircon fractionation which is largely controlled by temperature), whereas Y has similar properties as heavy rare earth element (HREE; albeit being present at higher abundances), and thus potentially can indicate garnet presence [e.g., Monteleone et al., 2007]. Each depth profile consisted of 100 cycles with 15 s count times. Following analysis, pit depths from the depth profiles were determined to be ∼15 µm using a KLA-Tencor MicroXAM® optical interferometer at UCLA. Data Processing The significantly longer duration for the depth profile (100 cycles) as compared to the spot-mode analysis (15 cycles) necessitates a different approach to correcting raw data for mass fractionation. Our approach for the depth profiles was to generate four separate depth-dependent mass fractionation factors (i.e., separate mass fractionation corrections for cycles 1–25, 26–50, 51–75, and 75–100). This is in contrast to the single mass fractionation correction applied to all 15 cycles making up the spot-mode analysis. For final calculation of isotopic ratios from the long depth profiles, individual cycles in each of the 25 cycle mass fractionation groups were then subdivided into blocks consisting of five cycles each. The result is that data from the depth profiles is reported as a series of 20 blocks each consisting of five cycles, whereas the data from the spot mode simply consists of all 15 cycles from that analysis. All of the U-Pb relative sensitivity corrections are derived relative to zircon standard AS3. In order to calculate trace element concentrations (Ti, Hf, and Y) for the depth profile analyses, individual cycles from each depth profile were grouped into blocks of five cycles, creating 20 blocks for each depth profile. This is to facilitate direct comparison between the trace element and U-Pb isotopic data for each depth profile. Because zircon standard AS3 is inhomogeneous with regard to trace element abundances, zircon standard 91,500 [Liu et al., 2010] was used to determine trace element relative sensitivity factors (RSF) relative to 94Zr2O+ assuming stoichiometric Zr abundances in standards and unknown. Data from a depth profile conducted on the 91,500 standard was broken down into 20 blocks, each consisting of five cycles. For each of these five cycle blocks the RSF factors for each element were determined by dividing the measured intensities from the standard by the known concentration of the element in the standard (Ti = 4.5 ± 0.4 ppm; Hf = 6250.0 ± 27.2 ppm; Y = 136 ± 2 ppm) [Liu et al., 2010]. These calculations generated 20 RSF factors (one for each five cycle block) as a function of depth that were then used to calculate concentration values for each of the corresponding five cycle blocks from the unknowns. The depth variation of the RSF factors was <4% for Ti, Hf, Th, and U, and ∼8% for AS3 zircons analyzed under the same conditions as the unknowns and yielded the following average (n = 4) concentrations: Ti = 27.4 ± 4.5 ppm, HfO2 = 1.37 ± 0.21 wt %, Y = 2300 ± 600 ppm, and U = 900 ± 600 ppm (Th/U = 0.85 ± 0.4; uncertainties are 1 standard deviation). Positive correlation exists between incompatible trace elements Hf-Y, Hf-U, Y-U, whereas Hf-Ti are negatively correlated. Although AS3 trace element abundances are heterogeneous, our values compare reasonably with previously reported trace element concentrations for AS3 zircons (Ti = 21.6 ppm, HfO2 = 1.20 wt %, U = 360 ppm, Th/U = 0.64) [Aikman, 2007; Black et al., 2004]. 4. Results The U-Pb isotopic ratios of young zircons are very sensitive to common lead corrections and applying a conventional 204Pb-based common lead correction to the 238U/206Pb isotopic ratios of the youngest (e.g., <5 Ma) zircons analyzed in this study introduces unacceptably large uncertainties. For this reason, all of the 238U/206Pb isotopic ratios of the zircons were only corrected for common lead using 207Pb, rationalizing that concordance can reasonably be presumed because of the young age and comparatively low U in the zircons [e.g., Baldwin and Ireland, 1995; Schmitt et al., 2003]. U-Pb data and ages are reported in a variety of ways. Table 2 summarizes Pliocene U-Pb ages determined in spot and depth-profiling modes when various zircon populations from each sample are grouped together (see below for explanation of selection criteria and age determinations). Table 3 reports U-Pb results for the spot-mode analyses, including zircons that yielded inherited ages not used in the age calculations reported in Table 1. Table 4 reports analytical results from the depth profiles including U-Pb isotopic data and Ti, Hf, and Y concentrations. For the depth profile analyses, Ti concentrations are used to calculate model zircon crystallization temperatures based on the calibration (at 1 GPa): T(°C) = (−1/((Log (Ticonc) − 5.711)/4800)) − 273.15 [Ferry and Watson, 2007]. The U-Pb data is also used in the construction of Figure 4, comparing the 206Pb/ 238U ages from the spot mode (shown here as probability density curves) with the depth profile analyses (shown here as age versus depth), and Figure 5, containing Tera-Wasserburg concordia diagrams for the young zircon populations determined in both analytical modes. Table 2. Summary of Ages Determined Using Several Approaches (Models) for the Young Zircon Overgrowths in Samples PNG 03-118m and PNG 06-21a, and All Zircons in PNG 08-10ga Tera-Wasserburg Regressions Weighted Means of Individual Ages Model 1: Upper Intercept Fixed at C.L. Model 2: No Fixed Upper Intercept Model 3: 206Pb/238U Ages Model 4: T.W. Lower Intercept Ages PNG 03-118m Young blocks of depth profiles (n = 38) 2.89 ± 0.29 Ma No regress