RESULTS AND INTERPRETATIONS

X-Ray Fluorescence Data

X-ray fluorescence (XRF) analyses conducted on board the JOIDES Resolution show that diabase and gabbro from footwall sites (1114 and 1117) and hanging wall sites (1109 and 1118) are chemically similar, varying only in K, Ba, Rb, and Sr (Taylor, Huchon, Klaus, et al., 1999). Movement of these mobile elements is likely associated with hydrothermal alteration, which is further supported by the presence of epidote and pyrite within veins of heavily altered samples. In addition, the abundance of these mobile elements increases as a function of the degree of alteration within the samples, as determined by thin section analysis and loss-on-ignition values, which tend to increase with increasing alteration. These data suggest, at least on a first-order basis, that mafic rocks from the various hanging wall and footwall sites are variably altered but petrogenetically related.

⁴⁰Ar/³⁹Ar Results

⁴⁰Ar/³⁹Ar analyses of diabase and gabbro samples were challenging because of the low K content within the plagioclase and pyroxene separates. An average of ~0.001%-0.006% K within the grains at all sites (Monteleone, 2000) rendered both total fusion and step heating experiments on clinopyroxene unsuccessful. As such, ⁴⁰Ar/³⁹Ar dating of clinopyroxene provided no interpretable results.

Although shipboard XRF analyses suggest that the diabase and gabbro rocks are petrogenetically related and vary only in amount of hydrothermal alteration, total fusion ages of plagioclase separates from Sites 1109, 1117, and 1118 are strikingly discordant and imprecise (Tables T1, T2). This apparent contradiction can be resolved by comparing ⁴⁰Ar/³⁹Ar apparent ages and variation of (⁴⁰Ar/³⁶Ar)_i (the ratio of the initial nonradiogenic Ar component) as revealed on an isochron diagram along with assessment of K alteration of plagioclase grains within each sample. Potassium alteration, observed in XRF whole-rock analyses, is also observed in plagioclase grains within the samples. Results from four selected diabase/gabbro samples from Cores 180-1109D-45R, 180-1117A-11R, 180-1117-alt, and 180-1118A-70R, indicate variable (⁴⁰Ar/³⁶Ar)_i ratios and a younging of ⁴⁰Ar/³⁹Ar apparent ages with increasing K alteration. Two end-member apparent ages can be identified and are interpreted to date plagioclase cooling and a younger K alteration event as discussed below.

Ophitic diabase from Site 1109 was shown to be the least altered by XRF and microprobe analyses (Monteleone, 2000). Diabase from Core 180-1109D-45R consists of ~50% plagioclase, ~50% clinopyroxene, and <1% pyrite, magnetite, and ilmenite as accessory phases. Microprobe point analyses and transects of plagioclase indicate percent K variations from 0.014 to 0.038 across individual grains (Monteleone, 2000). The ⁴⁰Ar/³⁹Ar step heat experiment yielded a saddle-shaped age spectrum with an imprecise minimum apparent age of 60.5 ± 10.7 Ma (Fig. F3). The analyses reveal a nonradiogenic trapped argon composition (⁴⁰Ar/³⁶Ar)_i of 323.8 ± 7.7, which is higher than the atmospheric ratio (295.5) and indicates the presence of either inherited or excess ⁴⁰Ar within the plagioclase grains. Figure F3 shows the corrected ⁴⁰Ar/³⁹Ar spectrum assuming an atmospheric (⁴⁰Ar/³⁶Ar)_i and the ⁴⁰Ar/³⁹Ar age spectrum assuming a (⁴⁰Ar/³⁶Ar)_i of 323.8 ± 7.7, derived from a best-fit line from an isochron diagram. A plot of (³⁹Ar)_K/(³⁷Ar)_Ca, proportional to the K/Ca ratio, shows little variation between steps, with a range of values from 1-2.5. Given the complex form of the spectrum, the isochron age of 58.9 ± 5.8 Ma is interpreted as the best estimate for the age of this plagioclase sample and we interpret the data to represent the timing of cooling following crystallization. The relatively large error is due to the low K content of the sample.

Diabase Core 180-1118A-70R is the most altered of the selected samples, containing equal amounts of plagioclase and clinopyroxene, with extensive chlorite alteration, accessory pyrite, ilmenite, titanite, and magnetite (Fig. F4). Pyroxene grains contain ~0.001% K, similar to those from Core 180-1109D-45R, but plagioclase is extensively altered. Microprobe point analyses show a random variation of K content within plagioclase grains, with K ranging from 0.01% (as in unaltered Core 180-1109-45R) to ~6.5% K in areas of abundant alteration (Monteleone, 2000). ⁴⁰Ar/³⁹Ar analysis of Core 180-1118A-70R thus provides an estimate for the age of the alteration event. ⁴⁰Ar/³⁹Ar step heat results for this sample yielded a complex spectrum without a well-defined plateau. Although extensively altered, the (⁴⁰Ar/³⁶Ar)_i is close to atmospheric in composition (300.8 ± 1.8) (Fig. F4). The (³⁹Ar)_K/(³⁷Ar)_Ca ratio is variable, ranging from 40 to 200, which is one to two orders of magnitude higher than (³⁹Ar)_K/(³⁷Ar)_Ca ratios for unaltered Core 180-1109D-45R. We infer that most of the argon extracted from this sample is due to outgassing of potassic alteration phases (e.g., sericite) within the plagioclase. The isochron age of 31.0 ± 0.9 Ma is interpreted as the maximum age of alteration. The higher K content within the altered plagioclase (Core 180-1118A-70R) allows for higher precision in ⁴⁰Ar/³⁹Ar apparent ages as compared to the low percent K content of unaltered plagioclase from Core 180-1109D-45R.

Ophitic gabbro Core 180-1117A-11R is variably altered, and thus it is difficult to interpret results in the context of a plagioclase cooling age or age of K alteration. The sample has a similar modal mineralogy to those from Sites 1109 and 1118 but with coarser grain size (Fig. F5). Two separates were prepared, each exhibiting different degrees of K alteration within plagioclase grains. These were separated magnetically, based on the assumption that the samples would have varying degrees of magnetic susceptibility as a function of the degree of alteration. Magnetic separation produced white grains (nonmagnetic and less altered) and green grains (magnetic and altered). The less altered plagioclase separate is referred to as Core 180-1117A-11R, and the more altered separate is referred to as Core 180-1117-alt. Microprobe analysis reveals that alteration is variable between individual plagioclase grains of Core 180-1117A-11R, with overall K alteration intermediate between pristine plagioclase of Core 180-1109D-45R and extensively altered Core 180-1118A-70R (Monteleone, 2000). A ⁴⁰Ar/³⁹Ar step heat experiment on Core 180-1117A-11R (less altered separate) yielded a saddle-shaped age spectrum (Fig. F5; Table T2). On an isochron plot, the data reveal an age of 43.6 ± 2.5 Ma, with a corresponding (⁴⁰Ar/³⁶Ar)_i of 355.4 ± 4.2. Although the isochron age cannot be related to a specific tectonic event, as expected given its moderate degree of K alteration, it is intermediate in age between isochron ages for unaltered Core 180-1109D-45R (58.9 ± 7.7 Ma) and pervasively altered Core 180-1118A-70R (31.0 ± 0.9 Ma). It likely reflects the outgassing of a mixture of unaltered plagioclase grains and moderately altered plagioclase grains. The (³⁹Ar)_K/(³⁷Ar)_Ca is variable and higher than corresponding (³⁹Ar)_K/(³⁷Ar)_Ca ratios for unaltered Core 180-1109D-45R.

Core 180-1117-alt also yielded a complex age spectrum (Fig. F6; Table T2). The data plotted on an isochron plot yield an age of 35.4 ± 1.0 Ma, which corresponds to a (⁴⁰Ar/³⁶Ar)_i of 392.3 ± 8.7. The isochron age for Core 180-1117-alt is slightly older than the isochron age obtained on the most altered sample (Core 180-1118A-70R) (31.0 ± 0.9 Ma).

U/Pb Results

A number of zircons were identified in diabase Core 180-1117A-11R, although many were too small for analysis. Seven analyses of zircon are presented in Table T3 and on a Tera-Wasserburg plot in Figure F7. These zircons show very high trace element concentrations with a range in U concentrations from 656 to 7609 ppm, Th from 2099 ppm to near 3 wt%, and Th/U from 1.93 to 6.95. U-Pb ages from these zircons are more scattered than would be predicted from a single population dispersing between radiogenic and common Pb end-members. Five of the seven analyses lie within the error of a best-fit line, indicating an age of 66.4 ± 1.5 Ma (2 _m; mean squared weighted deviates [MSWD] = 1.55). Of the other two analyses, number 6.1 is slightly lower than the mean but has a large error. Analysis number 5.1, on the other hand, at 75.6 ± 3.6 Ma (2 ), is significantly older than the other analyses and their mean. It is unlikely that this age represents inheritance in the diabase. Rather, this age is symptomatic of an artifact of the U-Pb calibration in the presence of extreme U and Th concentrations. In this case, the Th for this analysis represents nearly 3 wt%, whereas all but number 6.1, which shows the low age, are <1 wt%. For this reason, both analyses 5.1 and 6.1 have been excluded. The remaining data lie within error of each other, despite a range in Th concentration from 2000 to 9000 ppm. The correlation between high U-Pb age and abnormally high U or Th concentrations is likely a result of enhanced Pb⁺ production, relative to U⁺, from a matrix with elevated mean density. Such observations are not unusual in secondary ion mass spectrometry, where ion production can be strongly influenced by matrix parameters. The ²³⁸U/²⁰⁶Pb zircon age of 66.4 ± 1.5 Ma for diabase from Hole 1117A is interpreted as the age of crystallization.

Small (2-5 cm) igneous and metamorphic clasts were recovered from the top of Pliocene to Holocene synrift sedimentary sections at Site 1108 and in Holes 1110B and 1111A; several were selected for thermochronologic analysis. The clasts are angular, and their size suggests a proximal source.

⁴⁰Ar/³⁹Ar Results

Two K-feldspar-phyric rhyolite clasts sampled from Holes 1110B and 1111A contain 1-5 mm moderately sericitized K-feldspar grains within a felsic groundmass. Electron microprobe analyses indicated that the groundmass chemistry for these clasts is similar (Monteleone, 2000), suggesting that the clasts are petrogenetically related, although K-feldspar grains are variably altered to sericite. ⁴⁰Ar/³⁹Ar total fusion ages from K-feldspar grains from these samples yield ages of 12.5 ± 0.3 and 14.4 ± 0.6 Ma (Table T2) for Cores 180-1110B-3X and 180-1111A-16R, respectively, suggesting that K alteration may have partially reset the ⁴⁰Ar/³⁹Ar apparent ages in both samples. ⁴⁰Ar/³⁹Ar step heat analyses produce disturbed spectra that can be interpreted to reflect the effects of partial resetting of argon systematics as the result of a subsequent thermal event, slow cooling, and/or hydrothermal alteration (Figs. F8, F9). Preserved volcanic textures and the presence of sericite alteration within K-feldspar grains makes low-temperature hydrothermal alteration the most likely cause of resetting of argon systematics. Integrated ⁴⁰Ar/³⁹Ar apparent ages from Cores 180-1110B-3X and 180-1111A-16R are 11.9 ± 0.2 and 14.1 ± 0.3 Ma, respectively. These are concordant with total fusion ages on the same samples (Table T2). The highest temperature step of Core 180-1111A-16R K-feldspar is concordant with the ²³⁸U/²⁰⁶Pb age of 15.7 ± 0.3 Ma (see below), suggesting that sericite may not have affected the argon systematics within the most retentive sites within the phenocryst. The lowest temperature step for each sample gave an apparent age between 10 and 11 Ma, which we interpret is the result of hydrothermal alteration subsequent to eruption.

U/Pb Results

Rhyolite from Sites 1110 and 1111 were considered together because of the similarities in the zircons and results. The zircons are typically tabular, 50-100 µm long, and have aspect ratios of ~3:1-5:1. U-Th-Pb data are presented in Table T3; a Tera-Wasserburg plot is shown in Figure F10. The zircons show a range in U concentrations from 74 to 465 ppm (on average higher in Site 1111) and Th concentrations from 1 to 130 ppm. The Th/U ratios are typically <0.1 (13 of 16 analyses); of the remaining three analyses, one analysis is of an inherited core. This inherited core has a U-Pb age of 96 ± 6 Ma (2 ), indicating the presence of Cretaceous crust in the area during the formation of the rhyolites. The remaining analyses all lie within error of a mixing line between radiogenic Pb at 15.7 ± 0.4 Ma (2 _m; MSWD = 0.44) and common Pb of the same age (Cumming and Richards, 1975). Thus, the K-feldspar-phyric rhyolite clasts cored near (and possibly deposited from) the Moresby Seamount are interpreted to be petrogenetically related, crystallized at 15.7 ± 0.4 Ma, and variably hydrothermally altered at 10-11 Ma.

Core 180-1108B-6R contains a ~3-cm microgranite clast deposited in the Pleistocene and recovered from the top of synrift oceanic sediments. The sample contains ~70% albite, ~15% microcline, ~10% quartz, and 5% biotite, chlorite, and amphibole. Accessory phases include pyrite, apatite, zircon, and sphene. ⁴⁰Ar/³⁹Ar total fusion ages from biotite and "feldspar" (a mixture of albite and microcline) are 2.6 ± 0.4 and 2.2 ± 0.6 Ma, respectively (Table T2). Apatite fission track dating gave an age of 2.7 ± 1.2 Ma (2 ) (Table T4). Apatite fission track dating was difficult because of the poor quality of apatite crystals and low spontaneous track density. However, the fission track age is within the error limits of the more precise ⁴⁰Ar/³⁹Ar total fusion ages for biotite and feldspar.

Only two zircons were identified from microgranite from Site 1108. They were small (<50 µm) and equant. One grain is 550 Ma and has 116 ppm U and 78 ppm Th, whereas a young grain at 3.1 ± 0.1 Ma has 1448 ppm U and 664 ppm Th. The common Pb content of the young grain is quite low, with a measured ²⁰⁷Pb/²⁰⁶Pb of 0.072 being close to the radiogenic end-member, and, therefore, the age is rather insensitive to the common Pb composition used. This grain may therefore represent the magmatic age, although a single analysis is not statistically reliable on its own. However, the close similarity to the range in ⁴⁰Ar/³⁹Ar total fusion ages does lend support to this interpretation. These data suggest that the microgranite crystallized at 3.1 Ma, rapidly cooled by ~2.6 Ma, and we infer it was shallowly emplaced.