©2014 Society of Economic Geologists, Inc.Economic Geology, v. 109, pp. 2179–2206

Layered Hydrothermal Barite-Sulfide Mound Field, East Diamante Caldera, Mariana Volcanic Arc

James R. Hein,1,† Cornel E.J. de Ronde,2 Randolph A. Koski,3 Robert G. Ditchburn,2 Kira Mizell,1 Yoshihiko Tamura,4 Robert J. Stern,5 Tracey A. Conrad,1 Osamu Ishizuka,6 and Matthew I. Leybourne7

1 U.S. Geological Survey, 400 Natural Bridges Dr., Santa Cruz, California 95060 2 GNS Science, 1 Fairway Dr., Avalon, PO Box 30-368, Lower Hutt 5010, New Zealand

3 U.S. Geological Survey, 345 Middlefield Rd., Menlo Park, California 940254 Institute for Research on Earth Evolution (IFREE), Japan Agency for Marine-Earth Science and Technology (JAMSTEC),

Yokosuka 237-0061, Japan5 Geosciences Department, University of Texas at Dallas, 800 W. Campbell Rd., Richardson, Texas 75080

6 Institute of Geoscience, Geological Survey of Japan/AIST, Tskuba 305-8567, Japan7 ALS Minerals, 2103 Dollarton Highway, North Vancouver, British Columbia, Canada V7H 0A7

AbstractEast Diamante is a submarine volcano in the southern Mariana arc that is host to a complex caldera ~5 ×

10  km (elongated ENE-WSW) that is breached along its northern and southwestern sectors. A large field of barite-sulfide mounds was discovered in June 2009 and revisited in July 2010 with the R/V Natsushima, using the ROV Hyper-Dolphin. The mound field occurs on the northeast flank of a cluster of resurgent dacite domes in the central caldera, near an active black smoker vent field. A 40Ar/39Ar age of 20,000 ± 4000 years was obtained from a dacite sample. The mound field is aligned along a series of fractures and extends for more than 180 m east-west and >120 m north-south. Individual mounds are typically 1 to 3 m tall and 0.5 to 2 m wide, with lengths from about 3 to 8 m. The mounds are dominated by barite + sphalerite layers with the margins of each layer composed of barite with disseminated sulfides. Rare, inactive spires and chimneys sit atop some mounds and also occur as clusters away from the mounds. Iron and Mn oxides are currently forming small (<1-m diam, ~0.5-m tall) knolls on the top surface of some of the barite-sulfide mounds and may also drape their flanks. Both diffusely and focused fluids emanate from the small oxide knolls. Radiometric ages of the layered barite-sulfide mounds and chimneys vary from ~3,920 to 3,350 years. One layer, from an outcrop of 10- to 100-cm-thick Cu-rich layers, is notably younger with an age of 2,180 years. The Fe-Mn oxides were <5 years old at the time of collection in 2009.

Most mound, chimney, and layered outcrop samples are dominated by barite, silica, and sphalerite; other sulfides, in decreasing order of abundance, are galena, chalcopyrite, and rare pyrite. Anglesite, cerussite, and unidentified Pb oxychloride and Pb phosphate minerals occur as late-stage interstitial phases. The samples con-tain high Zn (up to 23 wt %), Pb (to 16 wt %), Ag (to 487 ppm), and Au (to 19 ppm) contents. Some layered out-crop samples are dominated by chalcopyrite resulting in ≤4.78 wt % Cu in a bulk sample (28 wt % for a single lens), with a mean of 0.28 wt % for other samples. Other significant metal enrichments are Sb (to 1,320 ppm), Cd (to 1,150 ppm), and Hg (to 55 ppm).

The East Diamante mound field has a unique set of characteristics compared to other hydrothermal sites in the Mariana arc and elsewhere. The geochemical differences may predominantly reflect the distribution of fractures and faults and consequently the rock/water ratio, temperature of the fluid in the upper parts of the circulation system, and extensive and prolonged mixing with seawater. The location of mineralization is con-trolled by fractures. Following resurgent doming within the caldera, mineralization resulted from focused flow along small segments of linear fractures rather than from a point source, typical of hydrothermal chimney fields. Based on the mineral assemblage, the maximum fluid temperatures were ~260°C, near the boiling point for the water depths of the mound field (367−406 m). Lateral fluid flow within the mounds precipitated interstitial sphalerite, silica, and Pb minerals within a network of barite with disseminated sulfides; silica was the final phase to precipitate. The current low-temperature precipitation of Fe and Mn oxides and silica may represent rejuvenation of the system.

IntroductionThe Mariana volcanic arc is the ~1,400-km-long southern half of the Izu-Bonin-Mariana arc (Fig. 1A). The arc con-sists of 76 volcanic edifices grouped into 60 volcanic centers of which 26 (20 of those being submarine) are known to be hydrothermally or volcanically active (Baker et al., 2008). Most of the hydrothermal systems in the Mariana arc are

sulfur rich, producing predominantly native sulfur and silica deposits (de Ronde et al., 2004; Baker et al., 2008; Embley et al., 2007; Resing et al., 2009; Butterfield et al., 2011) and more distal manganese-oxide deposits (Hein et al., 2008). East Diamante submarine volcano in the sourthern part of the Mariana arc is the only known site along the arc to host hydrothermal barite and sulfide precipitation. East Diamante is a silicic complex caldera ~5 ×10 km in size and elongated east-northeast−west-southwest, which is breached along its

0361-0128/14/4268/2179-28 2179

† Corresponding author: e-mail, jhein@usgs.gov

2180 HEIN ET AL.

E. Diamante

145°E 150°E140°E135°E

-600

-700

-800

-900

-500

-100

0

-40

0

-1100

-300

-1200

-1300

-200

-1400

-200

-500

-30

0

-300

-400

-500

-300

-400

-30

0

-500

-300

-500

-400

-100

0

-400

-300

-800

-700

-1400

-500

-300

-400

-400

-30

0

-300

-800

-600

-300

-500

-400

-200

-500

145°43'E145°42'E145°41'E145°40'E145°39'E145°38'E145°37'E

15°58'N

15°57'N

15°56'N

15°55'N

15°54'N

°°0 2.5

Kilometers

Mound FieldReference Sample R04

Saipan

Guam

20°N

10°N

A

B

10,000 8,000 6,000 4,000 2,000 0Depth (m)

700 km

China

JapanS. Korea

Philippines

Indonesia

Papua NewGuinea

Dacite Domes

Fig. 1. A. Shaded bathymetric map (contour interval 1,000 m) of the Mariana volcanic arc system with the location of the East Diamante caldera shown by the red box. The location of the Mariana arc is given on the inset map. B. Shaded bathym-etry of East Diamante caldera, which shows the location of Figure 2 (small pink box) and the location of a reference Fe-Mn oxide sample HPD1012-R04 site (small red dot) from the southeast dome area located outside the mound field (see Tables 1, 3). The central cluster of resurgent dacite domes is indicated (bathymetric maps from Mariana Bathymetric, Susan Merle, PMEL/NOAA, comp.).

BARITE-SULFIDE MOUND FIELD, E. DIAMANTE CALDERA, MARIANA VOLCANIC ARC 2181

northern and southwestern sectors (Stern et al., 2013; Fig. 1B). A cluster of resurgent dacite domes rise 200 to 300 m above the center of the caldera floor 600 m below sea level (Baker et al., 2008). Among all known Mariana hydrothermal sites, black smoker chimneys have only been found at East Diamante caldera (de Ronde et al., 2004). However, Cu sul-fides occur along the lower inner wall of West Rota caldera where they occur as orthogonal vein networks that may be stockwork mineralization (Hein et al., 2009).

Three hydrothermal fields hosting sulfates-sulfides have been found in East Diamante caldera, all of which occur on the northeast flank of the resurgent dome complex (Figs. 1B, 2). One field consists of active sulfide chimneys and another of inactive sulfide chimneys (de Ronde et al., 2004). The third field hosts inactive, layered, elongate, barite-sulfide mounds capped by small knolls of active, low-temperature Mn and Fe oxides (Fig. 3A-C). This type of mound system has not been previously described and is the subject of this paper. This large field of barite-sulfide mounds was discovered in June 2009 and revisited in July 2010, using JAMSTEC’s R/V Natsu-shima and the remotely operated vehicle (ROV) Hyper-Dol-phin. The mound field was sampled during Hyper-Dolphin dive HPD1012 on cruise NT09-08 in 2009 and during dives HPD1150, 1151, and 1153 in 2010 on cruise NT10-12 (App. 1).

The full extent of the East Diamante mound field has not yet been determined. However, it strikes for at least 180 m east-west and >120 m north-south and is aligned along a series of NE-SW− to NW-SE−trending fractures. The part of the field that contains most of the barite-sulfide structures covers >120 × ~30 m (Fig. 2). Three types of sulfate + sulfide structures occur in the mound field: (1) layered barite-sulfide mounds,

some with inactive chimneys that grew from their top surface and others with small active oxide knolls that are growing from the upper surface (Fig. 3A-C, 4A); (2) individual, small, inactive chimneys, or chimney clusters (Fig. 4B); and (3) a massive bedded and fractured outcrop of barite, chalcopyrite, and pyrite (Fig. 4C). In addition, four types of Fe-Mn oxide deposits occur in the mound field: (1) small, active, oxide knolls atop the sulfate-sulfide mounds (Fig. 3B,C); (2) actively forming oxides draped over the flanks of some mounds (Fig. 3A); (3) actively forming, porous, very friable, Fe-Mn oxide deposits on the seabed throughout the area; and (4) inac-tive(?), indurated, centimeter- to decimeter-thick Fe oxide-rich layers that make up the seabed over hundreds of square meter areas to form a cap rock (Fig. 4D).

Methods

Chemistry and mineralogy

Samples were ground to ≤75 µm using an agate mortar and pestle for chemical and mineralogical analyses. A Philips X-ray diffractometer (XRD) with graphite monochromator was operated at 40 kv and 45 mA. Step scans were run from 4° to 71° 2q using CuKa radiation. XRD digital data were analyzed using Philips X’Pert High Score software to identify peaks and mineral composition.

Chemical compositions were determined using the follow-ing methods. For most samples, the 10 major elements (Si, Al, Fe, Na, Mg, K, Ca, Ti, Mn, P) were analyzed using lith-ium metaborate fusion and analyzed by inductively coupled plasma atomic emission spectroscopy (ICP-AES); minor ele-ments were determined by four-acid digestion and analyzed by ICP-AES (Ba, Cr, Cu, Li, Mg, Ni, S, Sr, V, Zn, Zr) and ICP-mass spectrometry (ICP-MS; Ag, As, Be, Bi, Cd, Co, Cs, Ga, Hf, In, La, Mo, Pb, Rb, Sb, Sc, Se, Sn, Ta, Te, Th, Tl, U, W), and rare earth elements (REEs) by lithium metabo-rate fusion and analyzed by ICP-MS. For samples with high Ba, Zn, Cu, or Pb contents (>10%), minor elements and the REEs were determined using sodium peroxide fusion fol-lowed by nitric acid digestion and analyzed by ICP-AES and ICP-MS. All samples collected on cruise NT09-08 had major elements determined by X-ray fluorescence spectrometry (XRF) of borate fused disks, which is our preferred method for the major elements in samples with low metal contents. For all sulfide-sulfate samples collected on this cruise, instru-mental neutron activation (INAA) was used to determine Ag, As, Au, Br, Co, Cr, Eu, Hf, Sb, Sc, Se, Ta, W, and Yb; Zn was determined by sodium peroxide fusion and analyzed by ICP-AES. For all samples, Cl was determined by specific-ion electrode, Hg by cold-vapor analysis, Au by fire assay and ana-lyzed by ICP-AES, H2O− by gravimetric analysis, total S by induction furnace/infrared spectrophotometry (IR), and SO4

2– and elemental S by leach/furnace/IR. Duplicate analyses were performed on 10% of the samples and the average error was approximately ±1% for all techniques.

Pearson product moment correlation coefficient matrices were calculated for the chemical data, which is a measure of the strength of linear dependence between two variables. Sta-tistical significance is given at either a 99% or 95% confidence level (CL). Q-mode factor analysis was used to identify com-mon groups of elements referred to as factors. On the basis

600

500

500 R4

R03,R05,R06

C01,R21,R22

145˚40'54"145˚40'48"

15˚56'42"

15˚56'36"

15˚56'30"

C01,RC0 R

Mound Field Discoveredin 2009

Active Chimney Field

Extinct Chimney Field

145˚41'00"

R05

R06,R08R07

R03,R04,R05R06,R07

HPD1012NT09-08

NT10-12HPD1150HPD1151HPD1153

0 100

m

m

m

Fig. 2. Bathymetric map showing sample locations within the mound field. The approximate minimum extent of the mound field is indicated. Open fractures, lineations, en echelon and elongate structures are oriented approxi-mately north-south (although vary from NE-SW to NW-SE) along the main trend of western sample sites. Also shown are the locations of the nearby active chimney field discovered during Submarine Ring of Fire cruises in 2004 ROPOS and 2006 Jason II dives, and the inactive chimney field discov-ered at the end of dive HPD1012 of NT0908 cruise.

2182 HEIN ET AL.

of XRD mineralogy and element correlations, we determine each factor to represent a particular mineral in the samples and elements in that factor to be associated with or contained within that mineral. This links mineralogy that reflects envi-ronment conditions to element contents that in part reflect sources. For Q-mode factor analysis, each variable percent-age was scaled to the percent of the maximum value before the values were row-normalized and cosine-theta coefficients calculated. Factors were derived from orthogonal rotations of principal component eigenvectors using the Varimax method (Klovan and Imbrie, 1971). All communalities, an index of the efficiency of a reduced set of factors to account for the original variance, are ≥0.90.

Sulfur isotopes

A mixed barite-sulfide sample was combusted to SO2 under flowing oxygen. The evolved SO2 was oxidized to sulfate by passing the gases through Br2 water. BaCl2 solution was then added to precipitate BaSO4. The d34S value of the sul-fate produced from the sulfide component of the sample was determined by mass spectrometry. Both combusted and non-combusted samples were digested in hot, concentrated HCl

for several hours to dissolve oxides or sulfides and the acid decanted. The residual material, mostly barite, was washed, transferred to Teflon PFA beakers, and digested with hot HNO3 plus HF to dissolve silica and silicates. Although the barite was then sufficiently pure for d34S measurement, the finest particles were removed by suspending in water and decanting because isotopic exchange during acid digestion slightly decreases their d34S value. The cleanest barite crystals were handpicked under a binocular microscope. Samples were measured in duplicate in tin capsules with an equal amount of V2O5 on a EuroVector elemental analyzer connected to a GVI IsoPrime mass spectrometer. All results are averages and standard deviations of duplicates and are reported with respect to VCDT, normalized to internal standards R18742, R2268, and R2298 with accepted d34S values of −32, +3.3, and +8.6‰, respectively. The external precision of the instrument was better than 0.3‰ for d34S values.

Radiometric ages

Small Fe-Mn oxide knoll samples were dated at the National Isotope Center, GNS Science Limited, Lower Hutt, New Zea-land, using the activity ratio 228Th/228Ra. The 228Th and 228Ra

A

B CC

C

~1 m

~0.5 m

Draped Fe-Mn Oxides

Fig. 3. Seabed photographs from ROV dives HPD1150 and HPD1012. A. Composite panoramic view of three en echelon mounds, each with a subsidiary oxide knoll on top. The flank of the closest mound is draped with Mn oxides (arrow) (dive HPD1150). B. Oxide knoll venting low-temperature hydrothermal fluids with both focused and diffuse flow (dive HPD1012). C. Close-up of (B) showing focused flow shimmering fluid rising vertically (area seemingly out of focus) from the orange-brown vent.

BARITE-SULFIDE MOUND FIELD, E. DIAMANTE CALDERA, MARIANA VOLCANIC ARC 2183

activities were determined by measuring their decay prod-ucts 208Tl and 228Ac in a well-type, high-resolution gamma-ray detector. Calibration of the detector and determination of activity ratios using 228Ra-228Th disequilibrium are discussed in Ditchburn et al. (2012). The sulfates-sulfides were dated using the decrease in 226Ra/Ba due to radioactive decay since the onset of mineralization. For example, ages were calcu-lated for older layers of barite, sphalerite, and galena by com-paring their 226Ra/Ba values with those of the young seafloor massive sulfides samples. The reader is referred to Ditchburn et al. (2012) for a detailed discussion of the techniques used to date our samples.

Scanning electron microscopy (SEM) and petrography

Singly polished thin sections were used for reflected-light microscopy and SEM (with attached energy dispersive X-ray analyzer, EDAX) analyses. After petrography, the thin sec-tions were carbon coated for SEM analysis and imaged using backscatter and secondary electrons.

Results

Description of the mound field

Individual layered mounds are typically 1 to 3 m tall and 0.5 to 2 m wide, with lengths from ~3 to ≤8 m (Fig. 3A-C). The mounds are composed of layers containing both barite and sulfide—predominately sphalerite (Fig. 5A, B)—with each layer bounded top, bottom, and at its ends by a barite-dominant, millimeter- to 2-cm-thick, white margin contain-ing disseminated sulfides (Fig. 5C, D). Rare, inactive spires and chimneys cap some of the mounds (Figs. 4A, 5E); fallen chimneys were buried in hydrothermal oxides. Small clusters of chimneys <2 m tall sparsely populate the area (Figs. 4B, 5F), whereas those situated atop the mounds can be several meters tall and characteristically quite thick, about a third the chimney height (Fig. 4B, D). One such chimney (named the “Sumo chimney”) that fell from a mound and was buried in oxide sediment was collected by us and has been analyzed in detail (see below). This chimney is ~1 m tall and up to 35 cm

D

BA

C

1153-R05x

1153-R04

x1153-R06

1153-R07

1151-R07 ~ 0.5 m

~ 1 m

~ 0.5 m

~ 2 m

Fig. 4. Seabed photographs from ROV dives HPD1151 and HPD1153. A. In the distance, several mounds, one from which sample HPD1153-R06 was taken (cross) and which supports an inactive chimney (solid arrow) on top. Note cluster of small inactive chimneys in foreground (dashed arrow). B. Cluster of inactive chimneys from which the top of a spire was collected (HPD1153-R07) as shown by the line and arrow. C. Layered outcrop cut by fractures from which samples HPD1153R04 and R05 were collected. Sample R05 was collected at the “x,” whereas sample R04 was collected from the same outcrop just outside the upper left corner of view. Barnacles (bluish area) can be seen inhabiting the base of the outcrop where low-temperature fluids are discharging along a fracture that can be traced for at least 8 m. The area marked by the black box is shown in Figure 7. D. Sample HPD1151-R07 collected from an Fe oxide-rich cap-rock layer.

2184 HEIN ET AL.

H

A

D

X

X

X

B

X

X

C

F

E

G

Fig. 5. Photographs of cut faces of hand samples. Digital scale bar in each photograph represents 3 cm. A. Sulfide-sulfate layer, with high Zn (23 wt %), Pb (10%), Ag (358 ppm), Au (19 ppm); sample HPD1153-R06 (see Fig. 4A). B. Mound layer showing shades of gray with porous (dominant) to massive parts; crosses mark areas analyzed for chemical composition; sample HPD1153-R03 (App. 2). Zn content is highest in the dark gray, porous middle part (19 wt %), whereas Ba is highest in the porous, pale gray left part (33 wt %). C. Typical mound layer showing mottled texture of barite-sphalerite-silica and rim of barite; gray-white mottled layer contains Zn (19 wt %), Ag (386 ppm), Au (6.4 ppm), and Cu (1%); sample HPD1150-R05B. D. Typical mound layer with uniform texture of gray barite-sphalerite-silica and barite rind; sample HPD1012-R22. E. A small, inactive chimney, or knob from top surface of a mound, or possibly a composite chimney with two conduits; sample HPD1151-R05. F. Top of dead chimney, very high in Zn, Pb, Cd, Ag, and Hg contents; sample HPD1153-R07 (see Fig. 4B). G. Sample of layered outcrop showing two patches of chalcopyrite (marked by crosses), with high Cu (28 wt %), sample HPD1153-R05 (see Fig. 4C). H. Oxide rock layer from seafloor cap-rock outcrop; sample HPD1151-R07 (see Fig. 4D). Although many fluid conduits are open, no flow was seen coming from this rock layer. Open conduits, oriented parallel to seabed, are lined with or filled with Fe oxide and filled in places between conduits with Mn oxide.

BARITE-SULFIDE MOUND FIELD, E. DIAMANTE CALDERA, MARIANA VOLCANIC ARC 2185

thick and has a central conduit lined with yellow silica (Fig. 6A-D); the conduit is completely plugged in its upper part by barite and silica.

The most Cu rich samples from the mound field were col-lected from a single outcrop (HPD1153-R04, R05; see Fig. 2). This outcrop is at least 5 m thick, covers more than 100 m2, and consists of layers ranging from ~10 cm up to 1 m thick of barite-sulfide (Figs. 4C, 5G). Each layer has a rind domi-nated by barite; translucent bacterial mats drape the outcrop. The outcrop is cut by parallel fractures that break it up into large blocks. Low-temperature, focused venting of hydrother-mal fluids from the fractures supports a biological community dominated by bacteria and barnacles (Fig. 7); gray sulfides are exposed along some fractures.

Precipitation of Fe and Mn oxides and silica is currently forming small (<1 m diam, ~0.5 m tall) knolls atop some of the larger barite-sulfide mounds; these precipitates also drape the flanks of some mounds (Fig. 3A-C). Both diffuse (through the walls)- and focused (through a conduit)-flow fluids are evident from shimmering water venting from the small oxide knolls, which are laced with anastomosing channelways. Yellow-brown oxides line active conduits and completely fill other conduits.

By contrast, black oxides fill spaces between conduits. Fluid temperatures were not measured but are considered to be low based on the mineral assemblage being precipitated. Similar textures and channelways occur in Fe oxide and silica cap-rock layers that characterize parts of the field, although these chan-nelways have predominantly a horizontal orientation (Figs. 4D, 5H). Fe-Mn oxides and silica are the only hydrothermal precipitates that occur on the seabed around the margins of the field and cover a much larger area downslope from the main barite-sulfide structures than they do upslope.

Mineralogy

XRD mineralogy of the small Fe-Mn oxide knolls and oxide cap-rock layers shows predominantly birnessite in the Mn-rich areas and goethite in the Fe-rich areas of the samples (Table 1). Talc was identified in one sample. Based on the geochemical compositions (see below) and the presence of a broad hump on the X-ray diffractograms, we infer that amor-phous silica occurs in most samples. Minor sulfide and sulfate minerals occur in some oxide samples.

The layered mounds and chimneys are composed predomi-nantly of barite and sphalerite, with lesser amounts of galena

DB

CA

Fig. 6. Photos of Sumo chimney, sample HPD1150-R06, an inactive chimney composed of barite, silica, sphalerite, and galena. A. Sample being unloaded from the ROV Hyper-Dolphin; the dashed line indicates the cross section where the sample was split for preliminary examination. B. Cross section through the chimney along the dashed line shown in (A). This half was used for chemical analysis and was cut in half again for analysis along the thin dashed line. C. Vertical cross section through the central third of the chimney, which is the left half of the sample shown in (B). D. Horizontal and vertical cross sections through the central chimney; the yellow conduit lining is amorphous silica, colored by a trace of Fe; red-brown-colored barite surrounds the chimney. Digital scale bar in each photo represents 10 cm.

2186 HEIN ET AL.

and anglesite; abundant pyrite and chalcopyrite occur only in the layered outcrop samples (Table 2), with lesser amounts of these minerals in nearby small clusters of chimneys. In addition, X-ray amorphous silica occurs in nearly every barite-sphalerite sample.

Petrology (polished thin sections and SEM)

An assemblage of barite, low Fe (<2 wt % by EDAX) sphal-erite, galena, and silica characterizes most samples, consistent with XRD results. Sulfides and silica are interstitial to barite. Chalcopyrite and pyrite are abundant only in layered outcrop samples. In the mound and chimney samples, galena gener-ally occurs as granular rims projecting into cavities present within porous sphalerite aggregates (Fig. 8A). Some galena (Fig. 8B) is rich in Sb (determined by EDAX). Silica forms a coating on all of these minerals. Anglesite and other lower temperature Pb minerals, including cerussite and unidenti-fied Pb oxychloride and Pb phosphate minerals, formed late in the paragenesis, after sphalerite and galena (Fig. 8B, C). The abundance of barite, the wide variety of Pb minerals, and especially the paucity of pyrite and other Fe sulfides contrast with the mineralogy typical of many other seafloor massive sulfide deposits in the global ocean, especially those forming at mid-ocean ridges.

In the mound and chimney samples, minor pyrite (and to a lesser extent sphalerite) occurs as interstitial spheroids

with concentric growth rings (Fig. 8D). Some samples show replacement and infilling of microfossils; for example, fora-minifera filled and replaced by colloform pyrite and silica (Fig. 8E). Evidence of microbial activity may be indicated by microborings at the margins and along fractures in some crys-tals, especially galena (Fig. 8F). Similar borings are found in volcanic glass in ocean crustal rocks (e.g., Furnes et al., 2001; Staudigel et al., 2008; Berkenbosch et al., 2012). We do not see the silicified bacteria (filamentous and spherical) that are common in a wide range of low-temperature hydrothermal deposits (e.g., Hein et al., 1999; Jones et al., 2008).

Fe-Mn oxide chemistry

The chemical composition of five Fe-Mn oxide samples was determined, which include: (1) two samples from knolls that were venting shimmering water, one of those a bulk sample and one an indurated fragment isolated from the surround-ing friable matrix (Table 3); (2) two samples of oxide cap-rock layers exposed at the seabed; and (3) one sample of talus that blankets the entire area and buries the base of the mounds; the Sumo chimney was buried in this material and the sample analyzed was adhering to the chimney. Layered cap-rock sam-ple R03 was collected near a mound in the southern part of the field, whereas layered cap-rock sample R07 was collected from a location about 20 m from a mound in the northern part of the field (see Fig. 2). Sample R07 has higher Cu, Ni, Co, and V contents than R03, which in turn has higher Ag, Ba, Pb, Hg, and Mo (Table 3). Samples R03 and R06A9 are con-sidered to bracket the range of element compositions of the oxides-silica blanket found at the seabed around the margins of the field. The Fe/Mn value varies dramatically among the five oxide samples, from 0.9 to 310, although the samples typi-cal of the small knolls and seafloor oxides have Fe/Mn values = 0.9 to 1.3; those typical of the seabed cap-rock layers have values between 7.2 and 8.2.

Mn correlates with Sr and K at the 99% confidence level (CL) and has an inverse correlation with Fe at the 95% CL. Iron has a positive correlation only with Be. Silica covaries with Co, Cu, Hf, V, Zr, Y, and the REEs at 95% CL. Bar-ium, Zn, Pb, Cd, and associated sulfate-sulfide elements (see below) also covary with each other.

Sulfate-sulfide chemistry

Barium (n = 20), Si (n = 5), Zn (n = 3), and Cu (n = 2) domi-nate the 29 barite-sulfide samples analyzed in this study. Zinc or Si is the second most abundant element in 18 of the Ba-rich samples, whereas Pb is second most abundant in two of these samples (App. 2). The general decreasing order of abundance of these main elements is Ba (mean 29 wt %), Si (10%), Zn

Fig. 7. Barnacles thriving along a fracture venting low-temperature hydrothermal fluids. This fracture cuts the layered outcrop and is outlined by the box shown in Figure 4C.

Table 1. XRD Mineralogy of Fe-Mn Oxide Samples from Mound Field1

Sample no. Description Major Moderate Minor

HPD1012-R21A Venting subsidiary mound (bulk) Goethite, birnessite Pyrite?HPD1012-C01A Venting (solid rock from friable matrix) Goethite Birnessite, ramsdellite?HPD1150-R03 Layered deposit next to mound (bulk) Goethite Birnessite Cuprite? buserite?HPD1150-R06A9 Scraped-off lower chimney flank Birnessite, goethite Barite, talc?HPD1151-R07 Seabed rock layer (bulk) Goethite Birnessite Talc, pyrite?

1 Major >25%, moderate 5 to 25%, minor <5%

BARITE-SULFIDE MOUND FIELD, E. DIAMANTE CALDERA, MARIANA VOLCANIC ARC 2187

Tabl

e 2.

XR

D M

iner

alog

y of

Hyd

roth

erm

al S

ulfa

te-S

ulfid

e Sa

mpl

es fr

om M

ound

Fie

ld1

Sam

ple

no.

Des

crip

tion

Maj

or

Mod

erat

e M

inor

HPD

1012

-R22

A

Bul

k gr

ay la

yer

from

a m

ound

B

arite

Sp

hale

rite

, gal

ena

HPD

1012

-R22

B

Thi

ck w

hite

rim

on

R22

A

Bar

iteH

PD10

12-R

22C

D

iffer

ent g

ray

laye

r fr

om th

e m

ound

B

arite

Sp

hale

rite

, gal

ena

HPD

1012

-R22

D

Diff

eren

t gra

y la

yer

from

the

mou

nd

Bar

ite

Spha

leri

te, g

alen

aH

PD10

12-R

22E

D

iffer

ent g

ray

laye

r fr

om th

e m

ound

B

arite

Sp

hale

rite

, gal

ena

HPD

1012

-R22

F

Thi

n w

hite

rim

on

R22

E

Bar

ite

Spha

leri

te, g

alen

aH

PD10

12-R

22G

D

iffer

ent g

ray

laye

r fr

om th

e m

ound

B

arite

Sp

hale

rite

, gal

ena

HPD

1150

-R05

B1

Dar

k gr

ay la

yer

from

a d

iffer

ent m

ound

B

arite

, sph

aler

ite

Ang

lesi

te

Gal

ena,

cha

lcop

yrite

, pyr

ite?

HPD

1150

-R06

A1

Sum

o ch

imne

y, o

uter

hal

f of w

all,

cent

ral s

egm

ent o

f chi

mne

y B

arite

Sp

hale

rite

, am

orph

ous

Gal

ena,

pyr

ite?

HPD

1150

-R06

A2

Sum

o ch

imne

y, in

ner

half

of w

all,

cent

ral s

egm

ent o

f chi

mne

y A

mor

phou

s Sp

hale

rite

, bar

ite

Gal

ena,

cha

lcop

yrite

?, p

yrite

?H

PD11

50-R

06A

6 Su

mo

chim

ney,

inne

r ha

lf of

wal

l, op

posi

te s

ide

from

6A

2 A

mor

phou

s, b

arite

Sp

hale

rite

G

alen

a, p

yrite

?H

PD11

50-R

06A

7 Su

mo

chim

ney,

out

er w

all,

oppo

site

sid

e fr

om 6

A2

Bar

ite

Am

orph

ous,

sph

aler

ite, g

alen

a Py

rite

?H

PD11

50-R

06A

3 Su

mo

chim

ney,

inne

r th

ird

of w

all,

cent

ral c

him

ney,

90°

from

6A

2 A

mor

phou

s B

arite

, sph

aler

ite

Gal

ena,

cha

lcop

yrite

pyr

ite?

chal

coci

te?

HPD

1150

-R06

A4

Sum

o ch

imne

y, m

iddl

e th

ird

of w

all,

cent

ral c

him

ney

Am

orph

ous,

bar

ite

Spha

leri

te

Gal

ena,

pyr

ite?

HPD

1150

-R06

A5

Sum

o ch

imne

y, o

uter

thir

d of

wal

l, ce

ntra

l chi

mne

y B

arite

Pyri

te?

HPD

1150

-R06

A8

Sum

o ch

imne

y, o

uter

mos

t whi

te r

ind

Bar

ite

Am

orph

ous

HPD

1150

-R06

A10

Su

mo

chim

ney,

con

duit

wal

l lin

ing

Am

orph

ous

B

arite

, sul

fur?

HPD

1150

-R06

B1

Top

of S

umo

chim

ney,

hal

f wid

th

Bar

ite

Am

orph

ous

Gal

ena,

sph

aler

ite, p

yrite

? H

PD11

50-R

06B

2 To

p of

Sum

o ch

imne

y, o

ther

hal

f wid

th

Bar

ite

Am

orph

ous

Spha

leri

te, g

alen

a, p

yrite

?H

PD11

51-R

05

Smal

l dea

d ch

imne

y or

kno

b on

top

of m

ound

B

arite

, sph

aler

ite

Gal

ena,

am

orph

ous

Pyri

te?

HPD

1153

-R03

A

Lay

er fr

om m

ound

: mas

sive

dar

k-gr

ay p

art

Bar

ite (5

6%)

Spha

leri

te, a

mor

phou

s, g

alen

aH

PD11

53-R

03B

L

ayer

from

mou

nd: p

orou

s da

rk-g

ray

part

B

arite

(33%

), sp

hale

rite

A

mor

phou

s, g

alen

aH

PD11

53-R

03C

L

ayer

from

mou

nd: p

orou

s pa

le-g

ray

part

B

arite

(57%

) A

mor

phou

s, s

phal

erite

G

alen

aH

PD11

53-R

03F

T

hick

bar

ite r

im o

f mou

nd la

yer

Bar

ite (6

2%)

Am

orph

ous

Spha

leri

te, g

alen

aH

PD11

53-R

04

Fie

ld o

f lar

ge b

lock

s an

d th

ick

laye

rs, p

iece

of o

ne b

lock

A

mor

phou

s Py

rite

, cha

lcop

yrite

B

arite

, sph

aler

iteH

PD11

53-R

05A

F

ield

of l

arge

blo

cks

and

thic

k la

yers

, len

s fr

om a

noth

er b

lock

sam

ple

Cha

lcop

yrite

(81%

) Py

rite

B

arite

HPD

1153

-R05

B

Fie

ld o

f lar

ge b

lock

s an

d th

ick

laye

rs, h

ost f

or le

ns in

R05

A

Bar

ite

Am

orph

ous

Cha

lcop

yrite

, sph

aler

iteH

PD11

53-R

06

Lay

er fr

om m

ound

with

dea

d ch

imne

y on

top

Spha

leri

te

Bar

ite, g

alen

a, a

mor

phou

s C

halc

opyr

ite, a

ngel

site

HPD

1153

-R07

A

Chi

mne

y to

p fr

om c

lust

er o

f sm

all c

him

neys

, pal

e-gr

ay p

art

Bar

ite

Spha

leri

te, a

ngel

site

, gal

ena

Cha

lcop

yrite

, am

orph

ous

HPD

1153

-R07

B

Chi

mne

y to

p fr

om c

lust

er o

f sm

all c

him

neys

, dar

k-gr

ay p

art

Spha

leri

te, b

arite

A

mor

phou

s, a

ngel

site

, gal

ena

Cha

lcop

yrite

1 M

ajor

>25

%, m

oder

ate

5 to

25%

, min

or <

5%; a

mor

phou

s is

like

ly h

ydra

ted

silic

a ba

sed

on 2

-the

ta lo

catio

n of

the

broa

d hu

mp

2188 HEIN ET AL.

(7.5%), Pb (3.8%), Fe (1.7%), and Cu (0.28%), excluding the two Cu-rich samples. These bulk compositions are consis-tent with the presence of barite, silica, sphalerite, galena, and chalcopyrite, as described above. Several precious and rare metals are strongly enriched in these sulfate-sulfide samples, including (mean/maximum contents) Au (3.4/19  ppm), Hg (17/55  ppm), Ag (237/487 ppm), Cd (348/1,150 ppm), As (375/1,430 ppm), and Sb (426/1,320 ppm). More detailed discussions of geochemistry are presented below for mounds, chimneys, and layered outcrop samples in that order.

Mound samples from several locations were analyzed in this study. Seven layers (HPD1012-R22A-R22G; see Fig. 2)

from the upper and central part of a single mound show a narrow range of major element contents for the five dark gray layers, Ba (40−46 wt %), Zn (4−7 wt %), Pb (2.1−4.4 wt %), Si (1.9−5.0 wt %), and Fe (0.24−0.57 wt %; App. 2) and a wider spread for minor elements, As (170−942 ppm), Cd (98−238 ppm), Sb (321−903 ppm), and precious metals, Ag (201−487 ppm; the highest Ag measured), Au (1.2−2.9 ppm). The two white rind samples (R22B, R22F) have higher Ba, 47 to 50 wt % (equiv to 79−85 wt % BaSO4).

A single gray layer (R05B1) from another mound shows much higher contents of Zn (18.5 wt %), Pb (5.7 wt %), Fe (2.6 wt %), Cu (1.0 wt %), and Au (6.4 ppm); much less Ba,

gn

basph

ang

si sph

sph

gn

ang

ba py

gn

gn

sph

gn

pypy

gn

sph

50 um

si

200 um

50 um

200 um

200 um

50 um

gn

A B

C D

E F

Fig. 8. SEM backscatter images of sulfates-sulfides. A. Typical sulfide texture showing porous sphalerite (sph) overgrown by late-forming galena (gn) projecting into a cavity (sample HPD1153-R07A). B. Spheroidal sphalerite with interstitial white anglesite (ang); white dendritic Sb-rich galena in sphalerite; dark silica rims (arrow) around spheroids (sample HPD1153-R05B). C. Galena (white euhedral) and anglesite (pale gray) interstitial to sphalerite (medium gray; sample HPD1153-R07A). D. Pyrite (py) spheroids interstitial to barite (ba) and coated by silica (si); faint arcuate galena layers in pyrite (sample HPD1153-R03C). E. Colloform pyrite suggesting replacement of foraminifera (cf. Berkenbosch et al., 2012); dark silica layers in pyrite, medium gray is sphalerite, white is galena (sample HPD1153-R07A). F. Galena crystals containing pits and vermiform borings, possibly the result of microbial interactions on crystal surfaces and along fractures; other minerals are barite and sphalerite (sample HPD1012-R22).

BARITE-SULFIDE MOUND FIELD, E. DIAMANTE CALDERA, MARIANA VOLCANIC ARC 2189

Table 3. Chemistry of Mound Field Oxides Normalized to 0% H2O–; (R04A) from Southeastern Flank of Resurgent Dome for Comparison1,2

Hydrothermal Venting subsidiary Venting subsidiary Layered deposit Scraped off lower 10- to 20-cm-thickType sediment mound (bulk) mound (rock)3 adjacent to mound flank of chimney seafloor rock layerElements HPD1012-R04A HPD1012-R21A HPD1012-C01A HPD1150-R03 HPD1150-R06A9 HPD1151-R07

Fe (%) 34.0 32.0 55.8 44.1 22.8 41.1Mn 12.7 23.9 0.18 5.40 24.6 5.67Si 5.72 4.04 2.27 3.21 3.60 7.24Al 0.17 0.06 0.05 0.18 0.13 0.37Ca 0.73 0.42 0.18 0.32 1.09 1.19Mg 0.77 0.51 0.28 0.47 1.42 1.48Na 2.19 1.68 0.71 0.95 2.47 1.45K 0.71 0.85 0.03 0.22 0.96 0.31P 0.61 0.49 0.50 0.53 0.34 0.76S 0.19 0.14 0.10 0.13 0.48 0.36H2O– 11.7 16.6 3.70 5.49 37.6 19.0LOI 22.4 19.2 13.9 15.9 31.9 22.4Ag (ppm) 0.23 1.15 0.09 12.0 0.73 0.57As 1,495 1,415 768 1,466 565 1,169Ba 3,318 4,089 342 3,217 7,662 1,821Be 1.8 1.9 3.9 3.2 0.7 <0.1Bi 0.27 0.31 0.67 0.47 1.22 1.33Cd 1.17 0.88 0.40 0.47 12.8 3.19Co 3.3 1.6 0.62 3.1 9.6 54.5Cr 6 11 9 9 4 8Cu 134 41 9.7 121 191 625Ga 7.9 15.0 7.7Ge 0.5 0.4 0.5 Hf 0.06 <0.02 0.03 0.04 0.03 0.17Li 12 1 1 1 19 13Mo 501 675 376 355 1458 165Nb 0.2 <0.1 0.1 0.2 <0.1 0.5Ni 12.9 <0.5 <0.5 2.7 16.1 193Pb 202 155 253 444 3700 127Rb 3.9 3.4 0.3 1.3 4.5 4.2Sb 11.6 12.8 25.2 25.4 32.2 26.4Sc 1.4 1.2 1.6 0.5 0.3 1.2Sn 7.8 5.0 <0.3 3.6 2.1 2.4Sr 414 1451 64 350 1761 617Te <0.05 <0.05 <0.05 0.19 0.30 1.02Th 0.2 <0.1 <0.1 0.1 0.3 0.1Tl 13.1 20.4 0.37 11.2 42.0 16.2U 10.8 16.3 14.1 38.4 10.6 6.9V 196 106 187 206 154 994W 12.5 8.8 8.7 11.5 4.1 9.4Zn 470 582 222 653 5158 724Zr 4.2 2.2 3.2 4.2 2.8 15.6La (ppm) 2.6 2.6 2.0 3.0 8.7 5.7Ce 3.5 3.2 3.1 4.2 8.9 5.2Pr 0.49 0.43 0.51 0.57 0.85 1.12Nd 2.4 2.3 2.4 2.7 3.9 4.9Sm 0.6 0.7 0.9 0.9 1.1 1.3Eu 0.17 0.25 0.35 0.18Gd 0.75 1.04 1.28 1.38 1.50 2.48Tb 0.08 0.11 0.17 0.22 0.21 0.42Dy 0.66 0.95 1.37 1.32 1.18 3.15Y 6.3 7.7 10.3 7.7 8.1 35.1Ho 0.18 0.22 0.31 0.27 0.23 0.86Er 0.40 0.58 0.73 0.68 0.67 2.68Tm <0.05 0.06 0.10 0.11 0.08 0.42Yb 0.3 0.4 0.6 0.6 0.6 2.9Lu 0.07 0.06 0.09 0.09 0.11 0.52Au (ppb) 22 8 17 6 8Hg 173 287 129 6,403 74 233

1 Blank means not analyzed2 The following elements are at or below detection limits (ppm in parenthesis): Ti (100), Cs (5), In (0.02), Se (0.2), Ta (0.05)3 Solid rock isolated from poorly consolidated matrix and mud

2190 HEIN ET AL.

with Ag and Si within the same range as the series of layers from the first mound (App. 2). Three subparts (HPD1153-R03A-R03F) of a gray layer (in addition to its white rind) from a third mound were analyzed and show high variability within the gray part of this layer; for example, Ba (19−33  wt  %), Zn (5−19 wt %), and Pb (2−5 wt %). This gray layer has the highest measured Ga content (45 ppm), as well as high Ag (182−314 ppm), Au (0.8−3.4 ppm), and Cd (223−999 ppm; App. 2). A different dark gray layer (HPD1153-R06) from a fourth mound has the highest measured Zn content (23.1 wt %), Sb (1,320 ppm), Au (19 ppm), and second high-est Cd (1,120 ppm) and Hg (55 ppm). This layer also has the highest Fe (6.8 wt %) apart from the two Cu-rich samples.

Chimneys from two locations were analyzed, one in some detail (see Fig. 6A-D). The ~150-kg fallen Sumo chimney was collected next to a complex of mounds and it most likely fell from the top of the adjacent mound. Eleven samples (HPD1150-R06A1-R06B2; App. 2) were taken along four transects of the chimney as well as the <1- to 3-mm-thick, yellow chimney lining (Fig. 6B-D). The conduit lining has the highest Si (42 wt %), As (1,430 ppm), and Tl (163 ppm) con-tents of all the samples measured. Considering the abundance of barite and mass-balance calculations for S, the maximum amount of elemental S that could be coloring the silica con-duit lining is 0.4 wt %. However, traces of Fe can also color the silica (Mark Hannington, writ. commun, July 2013) and is the more likely explanation for the color based on the mineral assemblage and setting. The Sumo chimney samples generally show that Fe, Si, Zn, As, Cd, Cu, and Au contents increase from the outer wall (not including the barite coating) to the inner wall, whereas S, Ba, Pb, Ag, Sb, Sr, and the LREEs show a concomitant decrease. The main difference between the middle transects and the upper transect is the much higher Ba, Ag, and Sb and lower Zn, Cu, and Cd contents in the near chimney-top transect.

The top 10 cm of another chimney (HPD1153-R07) was analyzed, which was part of two clusters of 20- to 100-cm-tall chimneys, each growing up from a slab of sulfate-sulfide and draped by translucent bacterial mat. The sample was split into pale gray and a dark gray subsamples (R07A and R07B). These subsamples have the highest Pb (13−16 wt %), Ag (441−483 ppm), Cd (726−1,150 ppm), Cu (1.2 wt %, exclud-ing the two chalcopyrite-rich samples), and Hg (55 ppm) contents measured, as well as high Au (7−13 ppm). A small knob-like protuberance, that may be the base of a small chim-ney on the upper surface of a mound, has a similar composi-tion, but has somewhat lower Pb, Cu, Ag, and Hg contents (sample HPD1151-R05).

The layered outcrop was sampled from each of two, approx-imately 1-m-thick layers. Only a small portion of each massive layer was collected: 0.2 kg with sample HPD1153-R04, and 2.3 kg with HPD1153-R05. Sample R04 was analyzed in bulk, whereas R05 was split into analyses of a chalcopyrite-rich lens and the material around that lens. R04, R05A, and R05B have Cu contents of 4.78 wt %, 28.2%, and 0.52%, respectively, in addition to high contents of Fe (18.9 wt %, 19.1%, 0.68%), As (240−1,700 ppm), Hg (8−14 ppm), while the highest mea-sured Sr content (1.2%) occurs in sample R05B (App. 2). Gold is high in the most Cu rich sample, at 2.6 ppm. These are the only samples that contain large amounts of pyrite (~32 wt %)

and chalcopyrite (~14 wt % in the bulk sample and ~80 wt % in the lens). The bulk sample has the highest Fe, As, Co, In, and Mo contents in the entire dataset.

A Pearson product-moment correlation coefficient matrix (Table 4), which excludes the two chalcopyrite-rich samples and averages the Sumo chimney data, shows that Ba correlates with Na and the light REEs (except Ce and Pr) at the 99% CL (Table 4); and the correlation between Ba and Nd is excellent (Fig. 9A). With all the Sumo chimney data included, Ba also correlates with Sr; Ba is inversely correlated with Pb, Zn, Cd, Cu, Sb, Au, and Si (Table 4; Fig. 9C, D). Silica correlates with U (95% CL). Zn correlates with Cd, S, Ga, Au, Sb, Pb, Fe, Cu, Hg, and Ag at 99% CL (Table 4; Fig. 10A, B); Cd and Zn have a strong correlation indicating a Cd-rich sphalerite (Fig. 10A). Lead correlates with Hg, Sb, Cu, Cd, S, Au, Zn, Fe, As, Cu, and Ag at 99% CL (Fig. 10C, F); the correlation with Hg is very high (Fig. 10D). The strength of the correlation coefficients indicates that Ga and Au are most strongly associ-ated with Zn, whereas Hg and Cu are most strongly associated with Pb; Sb and Ag have similar coefficients with Zn and Pb, although SEM-EDAX studies show that the galena is Sb rich.

Q-mode factor analysis groups elements into factors that reflect different mineral phases, as determined by XRD. Because of their significantly different chemical compositions than all the other samples analyzed, the two Cu-rich samples were not included in the analysis, which is statistically required based on communalities for those two samples of <0.9. Three factors account for 90% of the data variance: 43% for factor 1, 28% for factor 2, and 19% for factor 3. Factor 1 contains Ba, S, Ag, La, and Ce and represents barite and associated ele-ments; La and Ce were chosen to represent the LREEs. Fac-tor 2 contains Zn, Pb, Ag, Sb, Cd, Ga, Au, Hg, Mo, Cu, S, As, Tl, and Fe and represents the combined Zn and Pb sulfides; Q-mode did not separate these two sulfide minerals and their associated elements. The third factor is composed of Si, U, Lu, and Mo and represents the silica phase; Lu was chosen to represent the heavy REEs.

Sulfur isotopes

The S isotope compositions of a barite-sphalerite pair from a mound layer have been determined. The barite has a d34S value of 20.5‰, which is nearly identical within error to pres-ent-day seawater sulfate (~21‰; e.g., Paytan et al., 1998). The sulfide from that layer has a d34S value of −2.5‰, which is within the range of Kermadec arc hydrothermal sphalerite samples (de Ronde et al., 2011) and Mariana arc chalcopyrite and sphalerite (C.E.J. de Ronde, unpub. data).

Radiometric ages

A sample of dacite (HPD1012-R07) from the mound field yielded a whole-rock 40Ar/39Ar age of 20,000 ± 4,000 years (Stern et al., 2013). Four samples from small oxide knolls were all less than 4.5 years old at the time of collection (June 2009), using the 228Th/228Ra method of Ditchburn et al. (2012; Table 5). The 226Ra/Ba ages of five layers from three mounds range from 3,620 to 3,920 years; four samples from three chimneys are trending to younger ages (3,350−3,690 yrs), although they overlap within error with the mound layer ages. The sample from the layered outcrop is much younger at 2,180 years old (Table 5). These samples from the mound field are entirely

BARITE-SULFIDE MOUND FIELD, E. DIAMANTE CALDERA, MARIANA VOLCANIC ARC 2191

Tabl

e 4.

Pea

rson

Pro

duct

Mom

ent C

orre

latio

n C

oeffi

cien

t Mat

rix

for

Bar

ite-S

ulfid

e Sa

mpl

es fr

om th

e E

ast D

iam

ante

Mou

nd F

ield

1

n  

Si

Al

Ca

Mg

Na

S

Ba

Pb

Zn

A

g A

s C

d C

u

18

Fe

−0.0

154

−0.5

693

−0.1

355

−0.4

080

−0.6

324

0.87

46

−0.8

242

0.69

07

0.70

69

0.15

79

0.49

28

0.72

76

0.67

0518

Si

−0.1

680

−0.4

714

−0.2

149

−0.3

647

−0.1

688

−0.3

313

−0.2

523

0.04

66

−0.3

667

−0.4

280

0.08

87

−0.2

446

18

Al

−0.1

047

0.34

13

0.77

46

−0.3

341

0.45

18

−0.3

709

−0.1

788

0.30

52

0.05

46

−0.3

429

−0.3

008

18

Ca

0.

6684

−0

.059

6 −0

.280

4 0.

3582

−0

.140

7 −0

.403

2 −0

.208

1 −0

.072

0 −0

.446

7 −0

.019

818

M

g

0.

3882

−0

.593

9 0.

5479

−0

.450

1 −0

.582

8 −0

.227

8 0.

0208

−0

.589

7 −0

.516

118

N

a

−0.4

408

0.63

86

−0.3

166

−0.3

746

0.29

25

0.09

41

−0.4

628

−0.4

390

18

S

−0.8

230

0.75

88

0.91

14

0.43

87

0.43

51

0.86

38

0.74

6818

B

a

−0

.705

8 −0

.882

8 −0

.267

9 −0

.294

7 −0

.885

7 −0

.661

318

Pb

0.

7083

0.

6194

0.

6587

0.

7924

0.

8068

18

Zn

0.

5107

0.

3060

0.

9297

0.

6483

17

Ag

0.49

57

0.44

67

0.45

0917

A

s

0.38

22

0.51

8018

C

d

0.

5941

n  

Ga

Mo

Sb

Sr

Tl

U

La

Ce

Pr

Nd

Sm

Au

18

Fe

0.51

12

−0.0

254

0.63

47

−0.2

565

0.68

64

−0.1

869

−0.4

715

−0.2

563

−0.0

242

−0.7

541

−0.5

912

0.81

8418

Si

0.

0417

−0

.357

8 −0

.292

2 −0

.130

4 0.

2666

0.

4933

−0

.404

3 −0

.337

7 −0

.332

4 −0

.328

2 0.

0215

−0

.237

118

A

l −0

.106

8 0.

4793

−0

.008

8 −0

.266

7 −0

.600

5 −0

.128

6 0.

0926

−0

.066

7 −0

.033

9 0.

3346

−0

.050

9 −0

.197

118

C

a −0

.275

1 0.

1681

−0

.374

9 0.

5401

−0

.192

0 0.

3595

0.

5687

0.

6350

0.

6505

0.

4961

0.

2893

−0

.203

718

M

g −0

.450

7 0.

1673

−0

.400

4 0.

1122

−0

.374

1 0.

5377

0.

4103

0.

2919

0.

3291

0.

5438

0.

2698

−0

.420

018

N

a −0

.390

1 0.

2346

−0

.105

3 −0

.293

7 −0

.590

4 −0

.288

0 0.

1864

−0

.060

3 −0

.115

5 0.

5036

0.

0847

−0

.354

918

S

0.

7490

0.

2349

0.

8165

−0

.373

9 0.

5231

−0

.522

1 −0

.569

5 −0

.365

6 −0

.140

5 −0

.804

7 −0

.751

0 0.

9227

18

Ba

−0

.743

9 −0

.014

7 −0

.666

1 0.

3613

−0

.543

7 0.

1564

0.

6713

0.

3967

0.

1587

0.

9378

0.

7190

−0

.772

718

Pb

0.

5164

0.

1358

0.

8166

−0

.308

1 0.

1901

−0

.461

1 −0

.269

5 −0

.075

6 0.

1079

−0

.639

8 −0

.652

3 0.

7338

18

Zn

0.91

01

0.32

17

0.81

98

−0.5

052

0.39

60

−0.4

918

−0.6

946

−0.4

730

−0.2

523

−0.8

933

−0.8

405

0.83

8317

A

g 0.

3765

0.

5412

0.

6773

−0

.509

3 −0

.212

3 −0

.661

3 −0

.192

7 −0

.134

4 0.

0141

−0

.266

3 −0

.639

1 0.

4072

17

As

0.09

53

0.09

44

0.68

05

−0.3

346

−0.0

020

−0.3

003

−0.0

402

−0.0

077

0.11

63

−0.2

951

−0.4

361

0.51

9818

C

d 0.

8234

0.

2223

0.

8664

−0

.500

8 0.

4235

−0

.389

4 −0

.654

4 −0

.482

2 −0

.299

3 −0

.919

5 −0

.758

4 0.

8454

18

Cu

0.50

39

0.02

98

0.59

18

0.07

82

0.12

44

−0.3

797

−0.1

346

0.16

05

0.32

93

−0.4

938

−0.4

843

0.68

0418

G

a

0.34

79

0.66

59

−0.3

331

0.27

00

−0.4

223

−0.5

282

−0.3

148

−0.1

711

−0.7

436

−0.6

698

0.65

7718

M

o

0.

5063

−0

.435

4 −0

.172

0 −0

.189

1 −0

.239

8 −0

.257

0 −0

.073

6 −0

.156

4 −0

.490

1 0.

5101

17

Sb

−0

.577

5 0.

2262

−0

.566

8 −0

.455

9 −0

.365

1 −0

.172

8 −0

.727

2 −0

.735

3 0.

8547

18

Sr

−0.1

675

0.33

28

0.68

70

0.74

63

0.54

62

0.55

55

0.74

82

−0.4

144

18

Tl

−0

.067

1 −0

.489

4 −0

.433

0 −0

.411

8 −0

.572

8 −0

.186

4 0.

3904

18

U

0.17

26

0.18

28

0.22

97

0.21

66

0.38

32

−0.3

573

18

La

0.

9205

0.

7389

0.

8378

0.

7135

−0

.584

818

C

e

0.

8946

0.

6523

0.

5392

−0

.400

918

Pr

0.44

06

0.21

05

−0.1

254

18

Nd

0.76

97

−0.8

012

18

Sm

−0

.742

4

1 C

u-ri

ch s

ampl

es 1

153R

04, 1

153R

05A

not

incl

uded

in th

e m

atri

x; S

umo

chim

ney

laye

r sa

mpl

es a

vera

ged

so a

s no

t to

bias

in fa

vor

of th

at c

ompo

sitio

n;n

= nu

mbe

r of

sam

ples

; zer

o po

int o

f cor

rela

tion

for

18 (1

7) s

ampl

es a

t 99%

con

fiden

ce le

vel (

CL

) is

0.58

2 (0

.597

), an

d at

95%

CL

is 0

.468

(0.4

82)

2192 HEIN ET AL.

different samples than those of the inactive East Diamante chimney field reported by Ditchburn et al. (2012).

DiscussionROV dives in 2009 and 2010 discovered a unique field of lay-

ered hydrothermal mounds with sparse distribution of chim-neys and a layered barite-sulfide outcrop that are rich in base, precious, and other metals (Fig. 11). Although hydrothermal mounds are found on other arc volcanoes such as the Ker-madec arc (de Ronde et al., 2011, 2014), their structures and/or compositions differ from those described here. The East Diamante mound field has many unique characteristics com-pared to other hydrothermal sites. For example, this mound field is the first one described from the Mariana arc, and those that occur elsewhere are composed predominantly of debris from collapsed chimneys that is cemented by relatively low temperature (<120°C) minerals precipitated from diffuse fluid flow (e.g., Herzig and Hannington, 2000; de Ronde et al., 2011). Most of those mounds are mineralogically and chemically zoned due to a process termed zone refining (e.g. Tivey et al., 1995; Koski et al., 2003), but this process has not

been documented as having occurred in the East Diamante mounds. East Diamante hosts a field of subparallel, elongate, layered mounds with their long dimension oriented along fractures, as documented from the ROV videos (e.g., Fig. 4C). Multiple layers compose the mounds, which are dominated by barite mixed with sulfides, and they include a wide variety of Pb minerals. By contrast, there is a paucity of pyrite, other Fe sulfides, and bulk Fe content in general, and elemental S, which further characterizes this unique system. In addition, much younger, currently active, roughly equal-dimensional Fe and Mn oxide knolls developed atop many of the layered mounds. Finally, no fossilized fauna have been collected from hydrothermal deposits in the mound field, a common occurrence elsewhere (e.g., Hannington et al., 1995; Juniper and Sarrazin, 1995). Iron sulfides are among the most com-mon minerals found in typical hydrothermal sulfide deposits whether formed at spreading centers or in arc systems (e.g., Herzig and Hannington, 2000; Halbach et al., 2003). The pau-city of elemental S in the mound field contrasts with all other known hydrothermal systems in the Mariana arc (de Ronde et al., 2004; Embley et al., 2007; Resing et al., 2007; Ditchburn

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0 20 40 60

Nd

(pp

m)

Ba (%)

0

5

10

15

20

25

30

35

0 20 40 60

Si (

%)

Ba (%)

0

5

10

15

20

25

0 10 20 30 40 50 60

Zn

(%)

Ba (%)

0

2000

4000

6000

8000

10000

12000

14000

0 20 40 60

Sr

(pp

m)

Ba (%)

0.92R = 2

n = 26 0.78R = 2

n = 19

0.44R = 2

n = 26

0.88R = 2

n = 16

0.91R = 2

n = 10

0.79R = 2

n = 7

A B

C D

Fig. 9. Scatter plots with Ba showing (A) an excellent positive correlation with Nd (also with other light and middle REE); (B) a moderately good correlation occurs with Sr using all data points (r2 = 0.38, n = 26; not shown); the data can also be viewed as two parallel regression lines, the lower one defined by all layers in a single mound (n = 7) and the upper one all the other data; however, the lower regression line is statistically significant only at the 95% confidence level (CL), whereas the combined dataset is significant at the 99% CL even though the coefficient is lower, the n value is much higher. Inverse correlations occur with (C) Si and (D) Zn (and also with Pb); the Zn-Ba data can be divided into two trends, one showing a more rapid increase in Zn with decreasing Ba; however, all the data combined is also statistically significant at the 99% CL.

BARITE-SULFIDE MOUND FIELD, E. DIAMANTE CALDERA, MARIANA VOLCANIC ARC 2193

et al., 2008; Butterfield et al., 2011). These chemical differ-ences likely result from different S and O fugacities in the mineralizing fluids and their subsurface fluid compositions.

The meter-scale layered barite-sulfide outcrop (loca-tion HPD1153-R03; Table 1) is also unique, not only to the Mariana arc, but elsewhere. This outcrop is younger than the mounds by 1,200 to 1,700 years and is relatively Cu rich and therefore the product of a later stage, higher temperature

(>250°C) hydrothermal mineralization. Only the exterior of two 1-m-thick beds were sampled from this outcrop. Thus, this structure warrants further investigation.

Iron and manganese oxides

The mound field is characterized by diffuse fluid flow that has occurred at one time or another throughout the entire field, producing Fe and Mn oxides that cover the base of the

0

200

400

600

800

1000

1200

1400

0 5 10 15 20 25

Cd

(pp

m)

Zn (%)

0 2000 4000 6000 8000

10000 12000 14000 16000 18000 20000

0 5 10 15 20 25

Au

(pp

b)

Zn (%)

0

200

400

600

800

1000

1200

1400

1600

0 5 10 15 20

Sb

(pp

m)

Pb (%)

0

10000

20000

30000

40000

50000

60000

0 5 10 15 20

Hg

(pp

b)

Pb (%)

0

200

400

600

800

1000

1200

0 5 10 15 20

As

(pp

m)

Pb (%)

0

2000

4000

6000

8000

10000

12000

14000

0 5 10 15 20

Cu

(pp

m)

Pb (%)

0.87R = 2

n = 26

0.68R = 2

n = 25

0.71R = 2

n = 250.85R =

2

n = 19

0.47R = 2

n = 250.65R =

2

n = 26

A B

C D

E F

Fig. 10. Scatter plots with Zn and Pb showing (A) an excellent positive correlation of Zn with Cd and (B) a positive cor-relation with Au. Positive correlations also exist between Pb and (C) Sb, (D) Hg, (E) As, and (F) Cu; all of these elements intercorrelate, but those displayed have their higher correlation coefficient with either Zn or Pb as shown; all correlations are valid at the 99% CL.

2194 HEIN ET AL.

mounds and fallen chimneys, and which coat the flanks of some mounds and the layered outcrop (Fig. 11). These oxides are distinctive in texture and chemical composition among hydrothermal oxides found elsewhere in volcanic arcs and mid-plate seamounts (e.g., Usui and Nishimura, 1992; Hein et al., 1997, 2008). For example, macroscopically distinct Fe and Mn oxide minerals occur in close proximity in the small knolls and seafloor cap-rock layers. This suggests that temper-ature and redox gradients must have varied rapidly over short distances within each deposit. This is atypical of hydrothermal strata-bound Mn deposits that are abundant in volcanic arcs, which show very strong fractionation of Fe (mean 0.89 wt %) from Mn (mean 47.5 wt %) for large deposits (e.g., Hein et al., 2008). Other distinguishing characteristics of the arc strata-bound and mound field oxides are the high Li (mean 515 ppm), Cd (24 ppm), Co (73 ppm), Ni (349 ppm), and Zn (3,058 ppm) contents in the strata-bound deposits relative to the mound area oxides, which have lower means (7 ppm Li, 4 ppm Cd, 14 ppm Co, 71 ppm Ni, and 1,468 ppm Zn; Table 3). By contrast, the mound area oxides have much higher mean Hg (1,425 vs. 18 ppm), As (1,077 vs. 68 ppm), Ba (3,426 vs. 1,508 ppm), and Pb (936 vs. 70 ppm) than the strata-bound Mn oxides (Table 3; Hein et al., 2008), which may reflect the abundance of volatile elements in the mineralizing fluids that would otherwise have been sequestered by the higher tem-perature phases.

The close spatial proximity of macroscopically pure hydro-thermal Fe and Mn oxide minerals in the mound field oxides also contrasts with hydrogenetic seamount Fe-Mn crusts. Bulk analyses of the mound field oxide samples appear similar

to hydrogenetic ferromanganese oxide deposits with subequal amounts of Fe and Mn, such as Fe-Mn crusts. However, the Fe and Mn minerals in hydrogenetic crusts are epitaxially intergrown, do not have macroscopically distinguishable Fe and Mn minerals, and the Mn oxide is Fe rich. In addition, seamount crusts are very rich in Co, Ni, Ti, Mo, REE, Bi, W, Tl, Te, Pt, and many other metals (Hein et al., 2013) com-pared to the mound field oxides. The high contents of many metals in crusts reflect in large part their very slow growth rates of only several millimeter/Ma (e.g., Hein et al., 2000).

The layered cap rock generally has open horizontal chan-nelways, which suggests that fluid flow was also horizontal through these layers. Cap-rock oxide sample R07 may cap a relatively higher temperature part of the mound field hydro-thermal system, resulting in higher Cu, Ni, Co, and V con-tents than in layered sample R03; R03 has higher Ag, Ba, Pb, Hg, and Mo (Table 3), elements typical of the barite-sulfide deposits. This is consistent with the cap-rock location in the northern part of the field, near the Cu-rich layered outcrop (Fig. 11).

Mound and layered outcrop structure

These elongate structures formed along fractures from mineralizing fluids that vented from short-linear conduits rather than from point-source conduits. Low-temperature fluids were seen venting along several meter-long segments of fractures during the 2010 ROV dives (Fig. 7). Barite formed a basic external structural framework and an internal porous network for the initial phase of mound growth, probably at temperatures of about 200° to 250°C (based on the mineral assemblage) during vigorous mixing with seawater (Fig. 12). As the permeability decreased and the hydrothermal fluids became insulated from the surrounding seawater, Zn and Pb sulfides precipitated within the barite framework network, likely at temperatures of about 230° to 260°C at the onset, and decreasing to about 180°C during the later stages (e.g., Hannington and Scott, 1988; Fouquet et al., 1993; Herzig et al., 1993). Fluids appear to have circulated throughout the entirety of the mounds, as do the diffusely flowing fluids today. During conductive cooling, silica precipitation coated all earlier precipitated phases and helped stabilize the mound structure. This process repeated itself several times for each mound creating multiple layers, each layer surrounded by barite with disseminated sulfides. Recurring heat-flux and enhanced fluid flow and circulation were essential in con-structing the mounds, which were likely associated with cracking of sealed conduits by overpressured fluids followed by fracture resealing due to mineralization within the fluid conduits (Fig. 11).

Barite and sulfides precipitated within the mound layers with little mineral zonation. Layers show various textures that reflect predominantly white barite, thoroughly mixed bar-ite and sulfide in uniformly gray layers, and patchy zones of barite and sulfides in mottled gray and white layers. These mottled layers are typically coarser grained than the uniformly gray layers. Porosities can vary significantly from layer to layer and reflect the availability of metals and permeability of the layers. Barite and silica commonly are more abundant toward the layer margins, indicating greater interaction (mixing) of the hydrothermal fluids with seawater.

Table 5. Ages for Mound Field Samples1

Sample Description

Fe-Mn oxide subsidiary mound2 228Th/228Ra Age (yrs)HD1012-R21 Pale red Fe oxide 0.89 ± 0.06 3.8 ± 0.4HD1012-R21 Dark brown-red Fe-Mn oxide 0.92 ± 0.03 4.0 ± 0.2HD1012-R21 Dark brown oxide channelway lining 0.92 ± 0.03 3.9 ± 0.2HD1012-R21 Pale red oxide channelway lining 0.99 ± 0.06 4.5 ± 0.5

Sulfate-sulfide samples3 226Ra/Ba Age (yrs)HD1012-R22 Mound layer 9.3 ± 0.4 3690 ± 110HD1012-R22-I Mound layer 8.4 ± 0.2 3920 ± 90HPD1150-R05B Mound layer 9.6 ± 0.3 3620 ± 90HPD1150-R06 Sumo chimney 9.3 ± 0.3 3690 ± 90HPD1151-R05 Small knob or chimney on mound 9.7 ± 0.2 3590 ± 80HPD1153-R05 Layered outcrop 17.9 ± 0.3 2180 ± 70HPD1153-R06 Mound layer 9.3 ± 0.3 3690 ± 90HPD1153-R06-I Mound layer 8.8 ± 0.4 3820 ± 120HPD1153-R07A Chimney 10.8 ± 0.2 3350 ± 80HPD1153-R07B Chimney 10.0 ± 0.2 3530 ± 70  Initial value4 45.9 ± 1.2

1 Isotope activities and ages are adjusted to the sample collection date and errors are 1 sigma

2 Ages calculated using 228Ra-228Th disequilibrium and relate to years elapsed since sample collection in June 2009

3 Ages calculated from the decrease in 226Ra/Ba value due to radioactive decay since mineralization

4 The initial 226Ra/Ba was determined by analyzing young active chimneys from the same proximity

BARITE-SULFIDE MOUND FIELD, E. DIAMANTE CALDERA, MARIANA VOLCANIC ARC 2195

Where buoyantly rising fluid breached the upper surface of the mounds, chimneys formed that have similar barite- and disseminated-sulfide margins and barite-sulfide-silica inte-riors as mound layers. Chemical zonation is more common in the few chimneys analyzed than in the mound layers (see above).

A similar sequence for the structural development of chim-neys and fault-fracture controlled fluid circulation has been proposed previously (e.g., Haymon, 1983; Fouquet et al., 1993; Herzig et al., 1993; Tivey, 2007). For mid-ocean ridge black smoker chimneys, anhydrite rather than barite typi-cally produces the initial framework; however, owing to its

retrograde solubility (Blount and Dickson, 1969), anhydrite rapidly dissolves. Without silica to provide insulation from the surrounding seawater and structural integrity, the chimneys eventually collapse into mounds of debris. Although both anhydrite- and barite-rich chimneys occur in arc settings, insulating and structurally reinforcing silica is not always pres-ent (e.g., Berkenbosch et al., 2012).

Inferred composition of the mineralizing fluids

The fluids from which the barite-sulfide structures precipi-tated are no longer venting in the mound field. Presently, Fe and Mn oxides and silica precipitate from low-temperature

Mound LayerFormation

Layered OutcropFormation

~20,000 yearsDacite Domes Formation

Magma

Fe-Mn Deposits Hydrothermal Magmatic Fluids Sea Water Fracture/Crack Sulfide Ore

Subsidiary Oxide KnollFormation

Mature Mounds Venting Fluid

Present Day

~ 3,920 - 3,620 years

~ 3,690 - 3,350 years

Layers Sealed via Mineralization

FractureBa, S, Zn, Pb, Cd, Sb, Hg...introduced via hydrothermal fluid (250-180 C)

Ba, Cu, Fe, Zn, Ni, Co, V, In...introduced via hydrothermal fluid

SO

Overpressuring causescracks and fissures, which allow the next

layer to form as heated fluids seep through

Sealed Barite Outer Layer

Occasional Chimney Formation

Block Faulting and Formation

of Outcrop

~ 2,180 years

< 2,180 years

Present Day

VolcanicBasement

Cap Rock

SubseafloorMineralization

1

2

24

4 3

5

Zn & Pb - RichMineralization

Pores/Vugs

Barite Matrix

4SO4SO4

SO4SO4

Fig. 11. Schematic model, not to scale, showing geologic setting and formation of mound field deposits: (1) dacite dome development occurred about 20,000 years ago, which was followed either by an absence of hydrothermal precipitation for 16,000 years, or the burial of earlier formed hydrothermal deposits; (2) hydrothermal layered barite-sulfide mounds formed during the period of about 4,000 to 3,600 years ago; and (3) sparse chimneys from about 3,700 to 3,350 years ago; forma-tion temperatures for the mounds and chimneys are inferred to have varied from ~250° to ~180°C; the hydrothermal fluids boiled at about 250°C; a host of elements (e.g., Ba, Zn, Si, Pb, Cd, Sb, Hg, Cu, Ag, Au) were derived from leaching of dacite basement, seawater, and to a lesser extent, magmatic fluids; (4) pyroclastic or volcaniclastic beds were mineralized subsurface ~2,180 years ago at the northern end of the field, which was later uplifted by volcanic-related growth faults or block faulting, which is seen now as a layered outcrop; the layered outcrop hosts a higher temperature metal assemblage (i.e., Cu, Ni, Co, In, V) than the mounds, metals that are also found in the lower temperature seabed cap-rock oxides that also occur in the northern part of the field and that may have acquired those metals from the leaching of subseafloor sulfide deposits; this type of Cu-rich layered deposits may have formed in the subsurface throughout the mound field area; (5) subsidiary Fe and Mn oxide knolls formed atop many of the barite-sulfide mounds and are less than a decade old; some of these knolls vent low-temperature (<120°C), clear fluids.

2196 HEIN ET AL.

diffuse-flow venting fluids, and more rarely from low-tem-perature focused-flow venting fluids. The hydrothermal fluids being discharged today may be diluted equivalents of those that formed the barite-sulfide structures. Unfortunately, as the fluids were not sampled, this cannot be verified. However, on the basis of the mineral assemblage and water depth, it can be inferred that fluids were below 260°C, had high sul-fur and variable oxygen fugacities, and experienced variable but widespread mixing with seawater. For the present water depth of the mound field, 367 to 406 m, boiling would have occurred at about 245° to 255°C (Bischoff and Rosenbauer, 1984, 1987). The fluids were also notable enriched in Ba, Zn, Pb, Si, Ag, As, Cd, Sb, Hg, and Au relative to other Mariana arc hydrothermal fluids (e.g., Resing et al., 2007). The sul-fide d34S value is more negative than the average value for sulfides from high-temperature mid-ocean ridge systems, but similar to those commonly measured for sulfides in arc systems (e.g., de Ronde et al., 2011, and references therein). The barite sulfate d34S value reflects the dominant seawater component of the fluids. Based on the mineral assemblage,

temperatures during most of the barite-sulfide precipitation were likely in the range of 180° to 250°C (see also Fouquet et al., 1993; Herzig et al., 1993). Fluid inclusion data for a simi-lar mineral assemblage from Axial Seamount and from two volcano hydrothermal systems of the Kermadec Arc are con-sistent with formation temperatures of ~250°C (Hannington and Scott, 1988; de Ronde et al., 2003a). Metals characteristic of higher temperature (>300°C) hydrothermal systems such as Sn, Se, W, Ni, Bi, Mo, and Co (e.g., Trefry et al., 1994; Herzig and Hannington, 2000) have very low contents in the mound field samples, except in the Cu-rich layered outcrop bulk sample (HPD1153-R04) with the highest Mo and Co contents. Boiling and phase separation most likely occurred subseafloor (cf. de Ronde et al., 2004) and could account for the high contents of volatile elements As, Sb, Hg, Ag, and Au. The final stages of mineralization probably saw an increase in fluid pH and precipitation of silica, which coats all the other minerals.

Source of the metals

The metals and other elements in the mound field deposits could have been derived from three sources and/or processes: (1) leaching of volcanic basement rocks, (2) magmatic fluids, and (3) seawater (Fig. 11). Metal contents in the deposits depend on the relative contributions from these sources, as well as composition of the host rocks, size and depth of the heat source, degree of leaching related to distribution of frac-tures, water/rock ratios, length of fluid pathways, and duration of fluid contact with host rock. Phase separation or boiling can also play a role in metal deposition (e.g., Fouquet et al., 1993; Halbach et al., 2003; Tivey, 2007). These factors con-trol or influence the composition, temperature, and pH of the hydrothermal fluid. Contributions of magmatic gases such as SO2, CO2, He, CH4, and H2 are common in volcanic arc set-tings (e.g., de Ronde, 1995; Yang and Scott, 1996; Gamo et al., 1997; Hannington et al., 2005; de Ronde et al., 2011). Leach-ing of silicic rocks such as rhyolite, andesite, and dacite results in vent fluids with high concentrations of many of the elements found in the mound field deposits, such as Ba, Zn, Pb, As, Sb, Cd, Au, Ag, and Si (e.g., Fouquet et al., 1993; Hannington et al., 2005; Stern et al., 2013). Considering the high contents of many of those metals in the mound deposits, it is likely that magmatic fluids also contributed to enrichments in those ele-ments with a magmatic affinity, for example, Au, Sb, As (e.g., Hedenquist and Lowenstern, 1994; Hannington et al., 2005; de Ronde et al., 2011). However, this same group of elements can be found in spreading ridge deposits that are considered by many workers to have little or no magmatic input, albeit in different ratios and typically with abundant Fe sulfides (e.g., Hannington et al., 1991). Ubiquitous Ba in the mound field deposits reflects pervasive leaching of feldspars in the high Ba source rocks (Stern et al., 2013) by the hydrothermal fluids.

Widespread Fe and Mn oxides precipitated from diffuse-flow fluids are common features of seafloor hydrothermal systems in their waning stages and also as the precipitate char-acteristic of the distal end of the seawater mixing during active discharge (e.g., Herzig and Hannington, 2000; de Ronde et al., 2003b). In the mound field, it is not known if these oxides (plus silica) were (1) deposited throughout the 2,000 years after cessation of sulfate-sulfide precipitation (waning stage),

Pb*10

Cu

Zn

ET

SS

MV

RS

OT

PL

PNMT

GB

sediment-freemid-ocean ridges

Juan deFuca R.

BS

(As+Sb)*100 Pb*10

Cu+Zn

sediment-freemid-ocean ridges

GB

OTETBS PL

PN

MT

Juan deFuca R.

MV

RS

Mound LayersChimneysLayered Outcrops

This paper, E. Diamante

TECTONIC SETTIN G

Back-arc basinArc volcanoRidge-hotspot

MT = Mariana TroughPL = Palinuro SeamountPN = Panarea SeamountOT = Okinawa Trough

Sedimentedenvironments

TECTONIC SETTIN G

From Hannington et al. (2005)

Fig. 12. Ternary plots modified from Hannington et al. (2005) and include data from the mound field barite-sulfide mound, chimney, and layered out-crop samples. Mound samples plot within the area occupied by data from the Palinuro and Panarea seamounts located in the Ionian Sea and the Mariana trough back-arc basin samples.

BARITE-SULFIDE MOUND FIELD, E. DIAMANTE CALDERA, MARIANA VOLCANIC ARC 2197

(2) represent a rejuvenation of the system, or (3) reflect distal precipitation contemporaneous with high-temperature min-eralization during an active stage that has predominantly been redirected away from the mound field. Only recent ages were determined for the oxides, which supports the rejuvenation scenario. However, because of the limited number of ages available, a final conclusion requires additional analyses and fluid sampling. The distal precipitation from high-tempera-ture, active-stage fluids is not likely to have occurred because they produce strongly fractionated Mn or Fe oxide deposits (Hein et al., 2008).

Some metal contents in the mound field samples are similar to those found in some seafloor massive sulfide deposits. For example, the Lau Basin deposits have high mean contents of Ag, Pb, and Cd and low Sn, Co, Se, Mo, and Ni when com-pared to the mound field samples. Similarly, Axial Seamount deposits in the northeast Pacific have similar high Au, Si, and Pb and low Cu, Co, Se, and Mo when compared to the mound field samples (e.g., Hannington et al., 1991; Fouquet et al., 1993; Herzig et al., 1993). However, the mound field samples have distinctive high mean contents of Ba and Sr and low Fe, Ga, In, Bi, and elemental S compared to other arc-system or mid-ocean ridge seafloor massive sulfide deposits (e.g., Hannington et al., 2005). Based on two ternary plots (Cu-Pbx10-Zn; Cu + Zn-As + Sbx100-Pbx10; Fig. 12) presented by Hannington et al. (2005), the mound field samples are compo-sitionally closest to Mediterranean Sea arc-system hydrother-mal deposits from Palinuro and Panarea seamounts (Ionian Sea), and those of the Mariana and Okinawa back arcs.

Temporal changes

A dacite sample from the mound field has been dated at ~20,000 years old, indicating relatively young volcanic activity along the east flank of the resurgent domes. By contrast, the oldest barite-sulfide deposit is ~4,000 years old, suggesting that still older deposits may have been covered by volcanic materials or gravity flows. Alternatively, there may have been a hiatus in hydrothermal activity after dome emplacement. Hydrothermal deposition in the mound field region spanned at least 4,000 years, during which the high-temperature vent-ing moved from the mound field to the dead chimney field and is now centered beneath the active chimney field (Fig. 2; de Ronde et al., 2004; Ditchburn et al., 2012). Within the mound field, high-temperature (~180°−260°C) mineraliza-tion spanned nearly 2,000 years, from ~4,000 to ~2,100 years ago, followed by low-temperature (<120°C) deposition. The low-temperature, small knoll hydrothermal oxides were less than 5 years old at time of collection in 2009; the age of the seafloor oxide cap-rock layers is unknown (Fig. 11).

The full geographic extent of the mound field and the min-eralization styles that it hosts are too poorly known to be spe-cific about the evolution of metal input into the system. Based on the data available, Cu-rich mineralization (i.e., the layered outcrop) occurred near the end of the period of relatively high temperature mineralization, paragenetically later than Zn- and precious metal-rich mineralization (Table 5).

On a small scale, the one chimney analyzed in detail (Sumo chimney) shows that generally Fe, Si, Zn, As, Cd, Cu, and Au increase systematically from the outer wall (excluding the barite outer layer) to the inner wall, whereas S, Ba, Pb, Ag,

Sb, Sr, and the LREE show the opposite trend. This indicates that with increasing isolation from seawater ingress, sphaler-ite (with Cd, As, Sb, and Au), silica (with As, Tl, and Hg), and possibly minor chalcopyrite precipitated. As the system waned and temperatures decreased, silica was the last phase to precipitate in lining of the conduit. Barite precipitated throughout the history of the chimney and during the waning stages was accompanied by galena (with Sr, Sb, Ag, and Au) and minor secondary Pb minerals. The metal distribution in the Sumo chimney is comparable to sphalerite-barite chim-neys found in Brothers volcano in the Kermadec arc, such as Lena chimney (de Ronde et al., 2011; Berkenbosch et al., 2012). However, notable Cu-rich mineralization seen in the Lena chimney does not occur in the Sumo chimney, which reflects the higher fluid temperatures (>300°C) involved in the formation of Lena (Berkenbosch et al., 2012).

Comparative geochemistry with nearby active and dead chimney fields

The active chimney field located near the mound field (Fig. 2) is among the shallowest black smoker systems known, with water depths around 345 m (de Ronde et al., 2004; Embley et al., 2007). On a broad scale, differences in the chimney and mound fields may reflect the distribution of fractures and faults and consequently the rock/water ratio, temperature of the fluid in the upper parts of the circulation system, and degree of mixing with seawater. Higher temperature chalcopyrite-pyrite-sphalerite deposits likely occur subsurface in the area of the mound field. By contrast, this style of mineralization occurs at the seabed as chimneys in the two chimney fields.

Compared to the mean chemical composition of chimneys in the two chimney fields (C.E.J. de Ronde, unpub. data), the mean composition of the mounds barite-sulfide samples shows an order of magnitude greater Au contents, as well as higher REEs (Tb is the same within error), Pb, Ba, and Si, and slightly more Ag. Conversely, the chimneys show two orders of magnitude more Ca, an order of magnitude more Cu and Ga, as well as higher Fe, Zn, Cd, As, S, and slightly more Hg and Sb than the mound samples (Table 6; Fig. 13). These dif-ferences reflect, in part, the more abundant anhydrite, pyrite, and chalcopyrite in the chimneys (C.E.J. de Ronde, unpub. data) formed by higher temperature fluids at the seabed that created the chimneys (measured at 242°C at 345-m water depth; Embley et al., 2007). The higher temperature fluids may in turn reflect a more direct route to the seafloor or lon-ger residence time of the fluid within the hot source rock, as well as decreased mixing with seawater, in contrast to the mound deposits which show much greater mixing with seawa-ter. The higher REE contents in the mounds reflect the higher barite contents, the principal host for the REEs. The much higher Au contents of the mound field deposits may reflect its late-stage incorporation into low-temperature sphaler-ite formed after prolonged mixing with seawater (Herzig et al., 1993). Galena and silica are part of that late-stage min-eral assemblage, but sphalerite is the dominant host for gold. This mineral association for Au is supported by the correlation coefficients and Q-mode factor analysis (see Table 4; Fig. 10).

The longer residence times of fluids in the mound struc-tures than in the chimneys and the greater surface area within mound layers may have helped to concentrate gold.

2198 HEIN ET AL.

Table 6. Chemistry Statistics for Mound Sulfates-Sulfide Samples, Excluding Two Cu-Rich Samples, and Mean Values for Nearby Active and Dead Chimney Field Samples (compiled from C.E.J. de Ronde, unpub. data)1,2

Mound field Mound field Mound field Mound field Mound field Mound field Chimney fields Chimney fieldsElements n Mean3 median minimum maximum st. dev. Mean4 n

Fe (%) 27 1.73 1.48 0.21 6.80 1.46 6.52 22Mn 27 0.02 0.01 0.01 0.12 0.02 0.13 13Si 27 10.0 5.10 1.39 41.6 10.2 3.04 13Al 27 0.08 0.07 0.02 0.19 0.05 0.19 13Ca 27 0.06 0.06 0.01 0.12 0.02 6.04 13Mg 27 0.03 0.02 0.01 0.10 0.02 1.43 13Na 27 0.46 0.09 0.02 2.05 0.67 0.26 22K 27 0.02 0.01 0.01 0.07 0.01 0.83 13Ti 27 0.03 0.01 0.01 0.10 0.03 0.60 13P 27 0.02 0.01 0.00 0.06 0.01 0.44 13Cl 19 0.11 0.09 0.03 0.33 0.09 -- --S 27 14.3 14 1.31 28.3 5.12 29.0 12Ba 27 29.4 33 3.61 49.8 13.5 6.68 22Pb 27 3.62 2.10 0.08 16.2 3.92 0.73 13Zn 27 7.48 5.76 0.21 23.1 6.57 26.2 22LOI 27 4.19 3.32 1.2 11.7 2.42 15.3 13Ag (ppm) 26 237 212 7.0 487 157 214 13As 26 375 232 60 1430 363 982 22Au 25 3.40 1.52 0.04 19.0 4.89 0.27 22Bi 27 0.64 0.2 0.2 2 0.72 0.40 13Br 6 7.13 6 4.8 14 3.48 19.5 22Cd 27 348 223 1.5 1,150 363 1123 13Co 26 4.96 1 0.5 31 8.38 65.6 22Cr 26 17.4 20 0.5 70 13.6 10.3 22Cu 27 2,849 1,630 110 12,000 3,148 49,605 13Ga 27 11.8 6 1.5 45 12.6 175 13Ge 7 5.14 4 2 9 2.61 8.77 13Hf 26 1.58 2 0.2 2 0.77 0.30 13Hg 20 16.8 12.1 0.75 54.9 15.8 17.6 22In 27 0.40 0.4 0.4 0.5 0.02 1.69 13Li 27 21.1 20 20 40 4.24 -- --Mo 27 41.1 37 3 110 26.8 119 13Nb 27 2.78 3 2 4 0.51 1.00 13Ni 27 10.4 10 10 20 1.92 3.13 13Rb 27 0.51 0.4 0.4 1.6 0.28 2.23 13Sb 26 426 313 75 1320 365 484 13Sn 27 3.30 2 2 17 3.66 1.00 13Sr 27 5,720 5,910 852 11,500 2,290 3,417 13Th 27 0.17 0.2 0.1 0.2 0.04 0.10 13Tl 27 22.4 10 0.75 163 33.2 15.4 13U 27 0.61 0.3 0.08 4.7 0.94 6.08 13V 27 12.6 10 10 40 6.56 37.6 13Zr 20 2.80 1.5 1 16 3.58 14.6 13La 27 9.27 10.2 0.5 16 4.37 1.06 13Ce 27 3.84 4 0.5 8.1 1.86 1.11 13Pr 27 0.18 0.16 0.1 0.4 0.09 0.11 13Nd 27 1.62 1.8 0.15 2.7 0.75 0.38 13Sm 27 0.84 0.9 0.15 1.6 0.39 0.13 13Eu 26 5.92 5 0.05 17.9 5.22 0.14 13Gd 27 0.26 0.1 0.1 0.88 0.27 0.12 13Tb 27 0.09 0.1 0.07 0.1 0.01 0.10 13Dy 27 0.09 0.1 0.05 0.1 0.02 0.11 13Y 27 1.30 1 1 2 0.47 2.00 13Er 27 0.09 0.1 0.05 0.1 0.02 0.10 13Yb 26 0.17 0.2 0.05 0.2 0.06 0.10 13Lu 27 0.22 0.08 0.05 1.6 0.34 0.04 13

1 Submarine Ring of Fire 2004 cruise to the Mariana arc, Ropos dives 787 and 7882 Submarine Ring of Fire 2006 cruise to the Mariana arc submarine volcanoes, Jason-2 dive J2-1933 All n values less than 27 represent no data and are not less than values; for Mn, Mg, K, Ti, P, Bi, Co, Cr, Hf, In, Li, Nb, Ni, Rb, Sn, Th, V, Zr, Pr, Eu, Gd,

Tb, Dy, Y, Er, Yb, and Lu, five or more samples reported less than values, for which the detection limit values were used to calculate the mean for those sam-ples; therefore, the true mean values for these elements are less than those listed in the table. For Pb, Ga, Mo, Tl, U, Ce, Nd, and Sm, four or fewer samples reported less than values, for which 75% of the detection limit was used to calculate the mean and are assumed to be good approximations

4 All n values less than 22 represent no data and are not less than values; for K, Ti, P, Ba, Bi, Br, Cr, Hf, In, Nb, Ni, Rb, Sn, Th, Tl, U, V, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Y, Er, Yb, Lu, and Au, three or more samples reported less than values, for which the detection limit values were used to calculate the mean, and therefore, the true mean values for these elements are less than those listed in the table; for Pb and Cu, greater than values were reported for one or more samples, for which maximum quantification limit values were used to calculate the mean, therefore, the true mean values are greater than those reported in the table. For Mn, Ga, Zr, and La, two or fewer samples reported less than values, for which 75% of the detection limit was used to calculate the mean and are assumed to be good approximations

BARITE-SULFIDE MOUND FIELD, E. DIAMANTE CALDERA, MARIANA VOLCANIC ARC 2199

There is no evidence for zone refining in the mounds depos-its, although it is common in other seafloor massive sulfide deposits (Koski et al., 2003); more detailed sampling by coring the mounds may show zone refining as a mechanism for the concentration of gold and other metals. However, unlike Au, the mean Ag content in the mound field deposits is similar to that in the chimneys. This may reflect the wider array of hosts for the Ag, which apparently precipitated with sphalerite and galena, and may form mineral inclusions in barite based on the Q-mode factor analysis. The much lower Fe in mound field deposits is not the result of different source rocks, as all three fields are hosted by dacite, which has lower Fe contents (3.2 wt %) than MORB (7.6 wt %; Qin and Humayun, 2008). Rather, the difference in Fe contents is due to much of the Fe in the mound field hydrothermal system being sequestered subsurface through precipitation of Fe sulfides and chalco-pyrite (Fig. 11). At present, Fe is sequestered in ubiquitous Fe oxide deposition throughout the mound field. The excep-tion is the layered outcrop, which shows precipitation of these higher temperature sulfide phases at the seafloor, or more likely exposed by faulting and uplift. This higher temperature mineral assemblage may be typical of subseafloor mineraliza-tion in the mound field.

SummaryA recently discovered field of barite-sulfide mounds capped

by small Fe-Mn oxide knolls occurs on the eastern flank of a cluster of resurgent dacite domes inside the East Diamante caldera, within sets of fractures that vary in orientation from northeast-southwest to northwest-southeast, based on ROV dive observations. This is the first mound field described from the Marina arc and is distinct in many ways (see “Discussion”

section). The mounds are elongate, layered structures that formed atop short, linear vents. A topographically distinct layered sulfate-sulfide outcrop occurs in the layered mound field and is unique within the Mariana arc. This outcrop is ~1,200 to 1,700 years younger than the ~3,350- to 3920-year-old mound layers and is a Cu-rich, inferred higher tempera-ture, but later stage feature (Fig. 11). The older mounds, and sparse chimneys, are barite-sphalerite-galena-silica−rich deposits, with the following significant mean (and maximum in brackets) trace metal enrichments: Au 3.4 (19) ppm, Ag 237 (487) ppm, and Cd 348 (1,150) ppm. The mound field deposits are also unique in their paucity of Fe sulfide minerals and the very low Fe contents in all samples, except those from the Cu-rich layered outcrop.

The mounds and layered outcrop were deposited during mixing of hydrothermal fluids with seawater at the seafloor with the precipitation of barite forming an external structural framework and internal porous network during the initial phase of mound growth. Then, with partial isolation from seawater, Zn and Pb sulfides precipitated within the bar-ite network (Fig. 11). During conductive cooling, a coating of silica precipitated on all earlier phases and helped form a stable mound structure. This process repeated several times for each mound creating multilayered structures, each layer surrounded by barite with disseminated sulfides. In order to create the layering in the mounds, reoccurring episodic enhanced fluid flow and circulation were essential and were likely associated with cracking and sealing of conduits. The cracking likely resulted from fluid overpressuring and the sealing from mineral precipitation.

Diffuse flow produced Fe and Mn oxides throughout the entire field that cover the base of the mounds and fallen

Fig. 13. Element enrichment diagram for mean composition of mound field sulfate-sulfide samples (excluding two Cu-rich samples) relative to mean composition of nearby chimney field samples (from Table 6; C.E.J. de Ronde, unpub. data); element ratios greater than 1 (above dark black line) are enriched in the mound field and those below 1 are enriched in the chimney field.

2200 HEIN ET AL.

chimneys and coat the flanks of some mounds and the layered outcrop. These oxides are chemically and texturally distinct from hydrothermal oxides found elsewhere in volcanic arcs and are also distinct from hydrogenetic crusts that form on mid-plate seamounts. Only the ages of the small knoll oxides are known, and these were less than 5 years old at the time of collection in 2009. These oxides were produced by low-temperature diffuse flow through the walls of mounds and knolls and also focused flow through vents in the small knolls. A unique feature of these oxide deposits is the close proximity of pure Fe and Mn oxide minerals, which form separate mac-roscopic phases that change abruptly over millimeters and reflect steep chemical gradients and rapid changes in pH and redox conditions. It is not known if these oxides (plus silica) were deposited throughout the 2,000-year period after cessa-tion of barite-sulfide precipitation, or whether they formed only in a late-stage rejuvenation of the system. However, we suggest that our age data for the oxides support the latter interpretation.

AcknowledgmentsWe thank the Captain and crew of the R/V Natsushima and

the ROV Hyper-Dolphin during cruises NT09-08 in 2009 and NT10-12 in 2010. We thank Francesca Spinardi and Kira Arias-La Rheir for technical assistance. Reviews by M.D. Hannington, J. Peter, D.A. Butterfield, and D.A. Clague were much appreciated and helped improve this contribution.

REFERENCESBaker, E.T., Embley, R.W., Walker, S.L., Resing, J.A., Lupton, J.E., Ko-

ichi, N., de Ronde, C.E.J., and Massoth, G.J., 2008, Hydrothermal activ-ity and volcano distribution along the Mariana arc: Journal of Geophysical Research, v. 113, no. B08S09, p. 1−16.

Berkenbosch, H.A., de Ronde, C.E.J., Gemmell, J.B., McNeill, A.W., and Goemann, K., 2012, Mineralogy and formation of black smoker chimneys from Brothers submarine volcano, Kermadec arc: Economic Geology, v. 107, p. 1613−1633.

Bischoff, J.L. and Rosenbauer, R.J., 1984, The critical point and two phase boundary of seawater, 200−500°C: Earth and Planetary Science Letters, v. 68, p. 172−180.

——1987, Phase separation in seafloor geothermal systems: An experimental study on the effects of metal transport: American Journal of Science, v. 287, p. 953−978.

Blount, C.W. and Dickson, F.W., 1969, The solubility of anhydrite (CaSO4) in NaCl-H2O from 100 to 450°C and 1 to 1000 bars: Geochimica et Cosmo-chimica Acta, v. 33, p. 227−245.

Butterfield, D.A., Nakamura, K., Takano, B., Lilley, M.D., Lupton, J.E., Res-ing, J.A. and Roe, K.K., 2011, High SO2 flux, sulfur accumulation and gas fractionation at an erupting submarine volcano: Geology, v. 39, p. 803−806.

de Ronde, C.E.J., 1995, Fluid chemistry and isotopic characteristics of sea-floor hydrothermal systems and associated VMS deposits: potential for magmatic contributions: Mineralogical Association of Canada Short Course Series 23, p. 479−509.

de Ronde, C.E.J., Faure, K., Bray, C.J., Chappell, D.A. and Wright, I.C., 2003a, Hydrothermal fluids associated with seafloor mineralization at two southern Kermadec arc volcanoes, offshore New Zealand: Mineralium Deposita, v. 38, p. 217−233.

de Ronde, C.E.J., Massoth, G.J., Baker, E.T. and Lupton, J.E., 2003b, Sub-marine hydrothermal venting related to volcanic arcs: Society of Economic Geologists Special Publication 10, p. 91−110.

de Ronde, C.E.J., Hein, J.R., Embley, R.W. and Stern, R.J., 2004, Hydrother-mal mineralization along the volcanically active Mariana arc [abs.]: EOS Transactions of the American Geophysical Union, v. 85(47), Fall Meeting Supplement, abstract V54A-04.

de Ronde, C.E.J., Massoth, G.J., Butterfield, D.A., Christenson, B.W., Ishi-bashi, J., Ditchburn, R.G., Hannington, M.D., Brathwaite, R.L., Lupton, J.E., Kamenetsky, V.S., Graham, I.J, Zellmer, G.F., Dziak, R.P., Embley,

R.W., Dekov, V.M., Munnik, F., Lahr, J., Evans, L.J., and Takai, K., 2011, Submarine hydrothermal activity and gold-rich mineralization at Broth-ers volcano, Kermadec arc, New Zealand: Mineralium Deposita, v. 46, p. 541−584.

de Ronde, C.E.J., Walker, S.L., Ditchburn, R.G., Caratori Tontini, F., Hannington, M.D., Merle, S.G., Timm, C., Handler, M.R., Wysoczanski, R.J., Dekov, V.M., Kamenov, G.D., Baker, E.T., Embley, R.W. Lupton, J.E. and Stoffers, P., 2014, The anatomy of a buried submarine hydrothermal system, Clark volcano, Kermadec arc, New Zealand: Economic Geology, v. 109, p. 2261–2292.

Ditchburn, R.G., de Ronde, C.E.J., Barry, B.J., Embley, R.W., Chadwick, W.W., Hein, J.R., and Butterfield, D.A., 2008, High-sulfidation submarine hydrothermal systems of the Mariana arc [abs.]: Underwater Mining Insti-tute, 38th, Oxford, Mississippi, Abstracts, p. 1−5.

Ditchburn, R.G., de Ronde, C.E.J., and Barry, B.J., 2012, Radiometric dating of volcanogenic massive sulfides and associated iron oxide crusts with an emphasis on 226Ra/Ba and 228Ra/226Ra in volcanic and hydrothermal pro-cesses at intraoceanic arcs: Economic Geology, v. 107, p.1635−1648.

Embley, R.W., Baker, E.T., Butterfield, D.A., Chadwick, W.W. Jr., Lupton, J.E., Resing, J.A., de Ronde, C.E.J., Nakamura, K.-I., Tunnicliffe, V., Dower, J.F., and Merle, S.G., 2007, Exploring the submarine ring of fire, Mariana arc, Western Pacific: Oceanography, v. 20, p. 68−79.

Fouquet, Y., von Stackelberg, U., Charlou, J.L., Erzinger, J., Herzig, P.M., Mühe, R., and Wiedicke, M., 1993, Metallogenesis in back-arc environ-ments: The Lau basin example: Economic Geology, v. 88, p. 2154−2181.

Furnes, H., Staudigel, H., Thorseth, I.H., Torsvik, T., Muehlenbachs, K., and Tumyr, O., 2001, Bioalteration of basaltic glass in the oceanic crust: Geo-chemistry, Geophysics, Geosystems, v. 2, paper 2000GC000150.

Gamo, T., Okamura, K., Charlou, J.-L., Urabe, T., Auzende, J.-M., Ishibashi, J., Shitashima, K., Chiba, H., and ManusFlux Scientific Party, 1997, Acidic and sulphate-rich hydrothermal fluids from the Manus back-arc basin, Papua New Guinea: Geology, v. 25, p. 139−142.

Halbach, P.E., Fouquet, Y., and Herzig, P., 2003, Mineralization and com-positional patterns in deep-sea hydrothermal systems, in Halbach, P.E., Tunnicliffe, V., and Hein, J.R., eds., Energy and mass transfer in marine hydrothermal systems: Berlin, Dahlem University Press, p. 85−122.

Hannington, M.D., and Scott, S.D., 1988, Mineralogy and geochemistry of a hydrothermal silica-sulfide-sulfate spire in the caldera of Axial Seamount, Juan de Fuca Ridge: Canadian Mineralogist, v. 26, p. 603−625.

Hannington, M., Herzig, P., Scott, S., Thompson, G., and Rona, P., 1991, Comparative mineralogy and geochemistry of gold-bearing sulfide deposits on the mid-ocean ridges: Marine Geology, v. 101, p. 217−248.

Hannington, M.D., Jonasson, I.R., Herzig, P.M., and Petersen, S., 1995, Physical and chemical processes of seafloor mineralization at mid-ocean ridges: Geophysical Monograph 91, p. 178−193.

Hannington, M.D., de Ronde, C.E.J., and Petersen, S., 2005, Sea-floor tec-tonics and submarine hydrothermal systems: Economic Geology 100th Anniversary Volume, p. 111−141.

Haymon, R., 1983, Growth history of hydrothermal black smoker chimneys: Nature, v. 301, p. 695−698.

Hedenquist, J.W., and Lowenstern, J.B., 1994, The role of magmas in the formation of hydrothermal ore deposits: Nature, v. 370, p. 519−527.

Hein, J.R., Koschinsky, A., Halbach, P., Manheim, F.T., Bau, M., Kang, J.-K., and Lubick, N., 1997, Iron and manganese oxide mineralization in the Pacific, in Nicholson, K., Hein, J.R., Bühn, B., and Dasgupta, S., eds., Manganese mineralization: Geochemistry and mineralogy of terrestrial and marine deposits: Geological Society, p. 123−138.

Hein, J.R., Koski, R.A., Embley, R., Reid, J., and Chang, S.-W., 1999, Diffuse-flow hydrothermal field in an oceanic fracture zone setting, northeast Pacific: Deposit composition: Exploration and Mining Geology, v. 8, p. 299−322.

Hein, J.R., Koschinsky, A., Bau, M., Manheim, F.T., Kang, J.-K., and Roberts, L., 2000, Cobalt-rich ferromanganese crusts in the Pacific, in Cronan, D.S., ed., Handbook of marine mineral deposits: Boca Raton, FL., CRC Press, p. 239−279.

Hein, J.R., Schulz, M.S., Dunham, R.E., Stern, R.J., and Bloomer S.H., 2008, Diffuse flow hydrothermal manganese mineralization along the active Mar-iana and southern Izu-Bonin arc system, western Pacific: Journal of Geo-physical Research, v. 113, no. B08S14, p. 1−29.

Hein, J.R., Stern, R.J., Tamura, Y., and Conrad, T., 2009, High temperature polymetallic sulfide feeder system, West Rota caldera, Mariana Islands: Deep-sea mining of seafloor massive sulfides: A reality for science and society in the 21st century: Science and Policy Workshop, 1−2 April 2009, Woods Hole, MA, Proceedings, p. 32−33.

BARITE-SULFIDE MOUND FIELD, E. DIAMANTE CALDERA, MARIANA VOLCANIC ARC 2201

Hein, J.R., Mizell, K., Koschinsky, A., and Conrad, T.A., 2013, Deep-ocean mineral deposits as a source of critical metals for high- and green-tech-nology applications: Comparison with land-based deposits: Ore Geology Reviews, v. 51, p. 1−14.

Herzig, P.M. and Hannington, M.D., 2000, Input from the deep: Hot vents and cold seeps, in Schulz, H.D., and Zabel, M., eds., Marine geochemistry: Berlin, Springer, p. 397−416.

Herzig, P.M., Hannington, M.D., Fouquet, Y., von Stackelberg, U., and Petersen, S., 1993, Gold-rich polymetallic sulfides from the Lau back arc and implications for the geochemistry of gold in sea-floor hydrothermal sys-tems of the southwest Pacific: Economic Geology, v. 88, p. 2182−2209.

Jones, B., de Ronde, C.E.J., and Renaut, R.W., 2008, Silicified filamentous microbes from the Giggenbach seamount: Implications for the interpreta-tion of fossil microbes: Journal of Geophysical Research, v. 113, B08S05, doi:10.1029/2007JB005482.

Juniper, S.K., and Sarrazin, J., 1995, Interaction of vent biota and hydrother-mal deposits: Present evidence and future experimentation: Geophysical Monograph 91, p. 178−193.

Klovan, J.E. and Imbrie, J., 1971, An algorithm and FORTRAN-IV program for large-scale Q-mode factor analysis and calculation of factor scores: Mathematical Geology, v. 3, p. 61−77.

Koski, R.A., German, C.R., and Hein, J.R., 2003, Mineralization and com-positional patterns in deep-sea hydrothermal systems in Halbach, P.E., Tunnicliffe, V., and Hein, J.R., eds., Energy and mass transfer in marine hydrothermal systems: Berlin, Dahlem University Press, p. 85−122.

Paytan, A., Kastner, M., Campbell, D., and Thiemens, M.H., 1998, Sul-fur isotopic composition of Cenozoic seawater sulfate: Science, v. 282, p. 1459−1462.

Qin, L. and Humayun, M., 2008, The Fe/Mn ratio in MORB and OIB deter-mined by ICP-MS: Geochimica et Cosmochimica Acta, v. 72, p. 1660−1677.

Resing, J.A., Lebon, G., Baker, E.T., Lupton, J.E., Embley, R.W., Massoth, G.J., Chadwick, W.W., and de Ronde, C.E.J., 2007, Venting of acid-sulfate fluids in a high-sulfidation setting at NW Rota-1 submarine volcano on the Mariana arc: Economic Geology, v. 102, p. 1047−1061.

Resing, J.A., Baker, E.T., Lupton, J.E., Walker, S.L., Butterfield, D.A., Mas-soth, G.J., and Nakamura, K.-I., 2009, Chemistry of hydrothermal plumes above submarine volcanoes of the Mariana arc: Geochemistry, Geophysics, Geosystems, v. 10, Q02009, doi: 10.1029/2008GC002141.

Staudigel, H., Furnes, H., McLoughlin, N., Banerjee, N.R., Connell, L.B., and Templeton, A., 2008, 3.5 billion years of glass bioalteration: Volcanic rocks as a basis for microbial life?: Earth-Science Reviews, v. 89, p. 156−176.

Stern, R.J., Tamura, Y., Ishizuka, O., Shukano, H., Bloomer, S.H., Emb-ley, R.W., Leybourne, M., Kawabata, H., Nunokawa, A., Nichols, A.R.L., Kohut, E., and Pujana, I., 2013, Volcanoes of the Diamante cross-chain: Evidence for a mid-crustal felsic magma body beneath the southern Izu-Bonin-Mariana arc: Geological Society London Special Publication 385.

Tivey, M.H., 2007, Generation of seafloor hydrothermal vent fluids and asso-ciated mineral deposits: Oceanography, v. 20, p. 50−65.

Tivey, M.H., Humphris, S.E., Thompson, G., Hannington, M.D., and Rona, P.A., 1995, Deducing patterns of fluid flow and mixing within the TAG active hydrothermal mound using mineralogical and geochemical data: Journal of Geophysical Research, v. 100, p. 12,527−12,555.

Trefry, J.H., Butterfield, D.B., Metz, S., Massoth, G.J., Trocine, R.P., and Feely, R.A., 1994, Trace metals in hydrothermal solutions from cleft segment on the southern Juan de Fuca Ridge: Journal of Geophysical Research, v. 99, p. 4925−4935.

Usui, A. and Nishimura, A., 1992, Submersible observations of hydrothermal manganese deposits on the Kaikata Seamount, Izu-Ogasawara (Bonin) arc: Marine Geology, v. 106, p. 203−216.

Yang, K. and Scott, S.D., 1996, Possible contribution of a metal-rich mag-matic fluid to a sea-floor hydrothermal system: Nature, v. 383, p. 420−423.

2202 HEIN ET AL.

Appendix 1. Sample Information

Sample Cruise Latitude Longitude Depth (m)

Sulfate-sulfide samples

HPD1012-R22A NT0908 15° 56.529 N 145° 40.927 E 377HPD1012-R22B NT0908 15° 56.529 N 145° 40.927 E 377HPD1012-R22C NT0908 15° 56.529 N 145° 40.927 E 377HPD1012-R22D NT0908 15° 56.529 N 145° 40.927 E 377HPD1012-R22E NT0908 15° 56.529 N 145° 40.927 E 377HPD1012-R22F NT0908 15° 56.529 N 145° 40.927 E 377HPD1012-R22G NT0908 15° 56.529 N 145° 40.927 E 377HPD1150-R05B1 NT1012 15° 56.53 N 145° 40.922 E 377HPD1150-R06A1 NT1012 15° 56.53 N 145° 40.920 E 377HPD1150-R06A2 NT1012 15° 56.53 N 145° 40.920 E 377HPD1150-R06A3 NT1012 15° 56.53 N 145° 40.920 E 377HPD1150-R06A4 NT1012 15° 56.53 N 145° 40.920 E 377HPD1150-R06A6 NT1012 15° 56.53 N 145° 40.920 E 377HPD1150-R06A7 NT1012 15° 56.53 N 145° 40.920 E 377HPD1150-R06A8 NT1012 15° 56.53 N 145° 40.920 E 377HPD1150-R06A10 NT1012 15° 56.53 N 145° 40.920 E 377HPD1150-R06B1 NT1012 15° 56.53 N 145° 40.920 E 377HPD1150-R06B2 NT1012 15° 56.53 N 145° 40.920 E 377HPD1151-R05 NT1012 15° 56.54 N 145° 40.920 E 371HPD1153-R03A NT1012 15° 56.55 N 145° 40.921 E 375HPD1153-R03B NT1012 15° 56.55 N 145° 40.921 E 375HPD1153-R03C NT1012 15° 56.55 N 145° 40.921 E 375HPD1153-R03F NT1012 15° 56.55 N 145° 40.921 E 375HPD1153-R04 NT1012 15° 56.55 N 145° 40.915 E 370HPD1153-R05A NT1012 15° 56.55 N 145° 40.915 E 370HPD1153-R05B NT1012 15° 56.55 N 145° 40.915 E 370HPD1153-R06 NT1012 15° 56.54 N 145° 40.916 E 371HPD1153-R07A NT1012 15° 56.54 N 145° 40.916 E 371HPD1153-R07B NT1012 15° 56.54 N 145° 40.916 E 371

Oxide samples

HPD1012-R04A NT0908 15° 55.857 N 145° 41.021 E 490HPD1012-R21A NT0908 15° 56.592 N 145° 41.927 E 377HPD1012-C01A NT0908 15° 56.529 N 145° 40.927 E 377HPD1150-R03 NT1012 15° 56.530 N 145° 40.922 E 376HPD1150-R06A9 NT1012 15° 56.530 N 145° 40.920 E 377HPD1151-R07 NT1012 15° 56.530 N 145° 40.980 E 377

Dacite samples

HPD1150-R05A NT1012 15° 56.53 N 145° 40.922 E 377HPD1151-R06 NT1012 15° 56.53 N 145° 40.93 E 381HPD1151-R08 NT1012 15° 56.52 N 145° 40.92 E 377

BARITE-SULFIDE MOUND FIELD, E. DIAMANTE CALDERA, MARIANA VOLCANIC ARC 2203

Appendix 2. Chemistry of Mound Field Sulfate-Sulfide Samples 1,2,3

Layer from Different layer Different layer Different layer Different layer mound: Layer from from same mound: from same mound: from same mound: Layer from from same mound: bulk central mound: white bulk central bulk central bulk central mound: white bulk centralType dark-gray part rim on R22A dark-gray part dark-gray part dark-gray part rim on R22E dark-gray part

Elements HPD1012-R22A HPD1012-R22B HPD1012-R22C HPD1012-R22D HPD1012-R22E HPD1012-R22F HPD1012-R22G

Fe (%) 0.33 0.93 0.57 0.36 0.24 0.29 0.33Mn 0.01 0.05 <0.01 <0.01 <0.01 0.02 0.03Si 4.67 3.98 4.96 2.65 1.92 1.39 4.00Al 0.14 0.06 0.11 0.13 0.12 0.15 0.19Ca 0.05 0.09 0.06 0.06 0.06 0.10 0.06Mg 0.02 0.05 0.02 0.02 0.02 0.05 0.05Na 1.85 0.65 1.31 1.61 1.97 0.76 2.05K <0.01 <0.01 <0.01 <0.01 <0.01 0.02 <0.01Ti 0.08 0.10 0.08 0.08 0.08 0.10 0.07P <0.01 0.01 <0.01 <0.01 <0.01 0.06 0.004Cl -- -- -- -- -- -- --S 12.9 11.5 13.6 14.4 15.1 12.0 12.6Ba 40.4 46.7 42.3 44.1 46.0 49.8 39.6(BaSO4) 68.7 79.4 71.9 74.9 78.2 84.6 67.3Pb 3.5 1.3 2.4 3.3 2.1 1.6 4.4(PbS) 4.0 1.5 2.8 3.8 2.4 1.8 5.1Zn 5.8 0.2 4.0 5.4 6.4 0.9 6.9(ZnS) 8.6 0.3 5.9 8.0 9.6 1.3 10.3LOI 2.92 2.10 2.71 2.43 2.63 2.00 3.32Ag (ppm) 211 201 487 295 379 -- 469As 441 511 224 170 376 -- 942Bi 0.7 1.8 2.3 1.1 2.2 1.8 2.1Br 6.9 14 5.1 5.2 6.8 -- 4.8Cd 197 4.0 98 135 169 21 238Co <0.5 <0.5 <0.5 <0.5 <0.5 -- <0.5Cr <0.5 <0.5 <0.5 <0.5 <0.5 -- <0.5Cu 1,420 450 1,430 1,180 1,480 910 2,660Ga 9 <2 5 7 8 7 14Ge 4 <2 4 6 8 3 9Hf <0.2 <0.2 <0.2 <0.2 <0.2 -- <0.2In <0.4 <0.4 <0.4 <0.4 <0.4 <0.4 <0.4Li <20 <20 <20 <20 <20 <20 <20Mo 50 50 40 60 60 110 60Nb <2 <2 <2 <2 <2 <2 <2Ni <10 <10 <10 10 <10 20 <10Rb 0.7 0.6 0.5 0.6 1.2 1.6 0.6Sb 354 163 321 410 558 -- 903Sn <2 <2 17 12 10 4 <2Sr 4,950 6,400 5,910 5,460 5,460 7,110 4,690Th <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1Tl 4 9 13 7 4 3 3U 0.10 0.64 0.13 0.10 0.08 0.70 0.25V 10 <10 <10 <10 <10 40 20Zr -- -- -- -- -- -- --La (ppm) 9.3 10.8 10.3 11.5 11.8 16.4 11.4Ce 3.3 4.3 3.6 4.0 4.3 6.5 4.0Pr 0.11 0.15 0.15 0.21 0.17 0.32 0.16Nd 2.0 2.3 2.4 2.2 2.5 2.7 2.0Sm 0.9 1.0 0.9 0.9 1.0 1.1 0.9Eu <0.05 <0.05 <0.05 <0.05 <0.05 - <0.05Gd 0.55 0.65 0.61 0.72 0.76 0.88 0.72Tb 0.07 0.07 0.07 0.08 0.07 0.09 0.07Dy <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05Y <5 <5 <5 <5 <5 <5 <5Er <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05Yb <0.05 <0.05 <0.05 <0.05 <0.05 - <0.05Lu 0.08 0.07 0.14 0.07 0.09 0.08 <0.05Au (ppb) 1,630 49 150 1,170 2,120 - 2,940Hg -- -- -- -- -- -- --

2204 HEIN ET AL.

Appendix 2. (Cont.)

Layer from Sumo chimney, Sumo chimney, Sumo chimney, Sumo chimney, a different Sumo chimney, Sumo chimney, inner third of wall, middle third of inner half of wall, outer wall, mound: bulk outer half of wall, inner half of wall, central chimney, wall, central opposite side opposite sideType central part central chimney central chimney 90o from 6A2 chimney from 6A2 from 6A2

Elements HPD1150-R05B1 HPD1150-R06A1 HPD1150-R06A2 HPD1150-R06A3 HPD1150-R06A4 HPD1150-R06A6 HPD1150-R06A7

Fe (%) 2.62 1.17 1.76 1.68 1.01 2.03 1.11Mn <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01Si 4.52 19.9 31.9 24.1 19.7 23.9 10.6Al 0.106 0.127 0.095 0.085 0.058 0.101 0.169Ca 0.09 0.04 0.01 0.03 0.04 0.02 0.05Mg 0.01 0.02 0.03 0.02 0.03 0.02 0.03Na 0.08 0.10 0.08 0.07 0.07 0.09 0.16K 0.02 0.02 0.02 0.01 0.01 0.01 0.04Ti <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01P 0.04 0.02 0.02 0.01 0.01 0.01 0.02Cl 0.06 0.13 0.1 0.09 0.07 0.11 0.24S 20.8 10.4 8.65 11.4 10.5 10.4 13.5Ba 22.1 22.1 3.89 12.8 19.0 16.3 32.2(BaSO4) 37.6 37.6 6.61 21.8 32.3 27.7 54.7Pb 5.7 1.6 0.5 1.1 2.1 0.9 3.5(PbS) 6.6 1.8 0.6 1.3 2.4 1.0 4.0Zn 18.5 5.5 9.6 10.4 5.2 6.5 6.4(ZnS) 27.6 8.2 14.3 15.5 7.8 9.7 9.5LOI 7.25 3.97 6.58 6.03 3.60 5.16 2.99Ag (ppm) 386 99 64 76 109 85 189As 170 60 90 60 80 110 90Bi <0.2 0.3 <0.2 0.4 0.2 <0.2 <0.2Br -- -- -- -- -- -- --Cd 427 247 629 660 210 393 257Co <1 12 31 24 14 19 12Cr <20 <20 <20 <20 <20 70 <20Cu 10,200 2,080 2,830 2,270 920 1,920 2,200Ga 38 6 6 8 2 5 4Ge -- -- -- -- -- -- --Hf <2 <2 <2 <2 <2 <2 <2In <0.4 <0.4 <0.4 <0.4 <0.4 <0.4 <0.4Li <20 <20 <20 30 40 <20 <20Mo 77 31 34 25 25 21 56Nb 3 3 3 3 3 3 3Ni <10 <10 <10 <10 <10 10 <10Rb <0.4 <0.4 <0.4 <0.4 <0.4 <0.4 <0.4Sb 456 101 85 88 152 96 231Sn <2 <2 <2 <2 <2 <2 <2Sr 7,690 5,200 1,110 3,250 4,000 4,240 6,030Th <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2Tl 17 2 5 2 <1 4 3U 0.1 1.2 2 0.3 4.7 0.3 0.5V 10 20 10 20 <10 20 <10Zr 2 <1 <1 <1 2 <1 2La (ppm) 9.4 7.1 0.7 3.7 6.7 4.4 9.9Ce 5.7 3.2 <0.2 1.1 2.5 1.4 4.2Pr 0.3 0.2 <0.1 <0.1 0.1 <0.1 0.2Nd 1.6 1.5 <0.2 0.7 1 0.8 1.7Sm 0.6 0.7 <0.2 0.3 0.6 0.5 1.1Eu 4.9 4.4 0.9 2.9 5.1 4.1 12.5Gd <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1Tb <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1Dy <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1Y 1 1 <1 <1 1 <1 2Er <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1Yb <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2Lu 0.2 <0.1 0.4 <0.1 0.3 1.6 1Au (ppb) 6,370 1,210 1,790 1,740 938 1,520 481Hg 16,100 5,370 5,790 5,580 5,350 4,730 6,280

BARITE-SULFIDE MOUND FIELD, E. DIAMANTE CALDERA, MARIANA VOLCANIC ARC 2205

Appendix 2. (Cont.)

Sumo chimney, Sumo chimney, Top of Sumo Top of Sumo Small dead chimney Layer from Layer from outermost conduit chimney, chimney, other or knob on top mound: massive mound: porousType white rind wall lining half width half width layer of mound dark-gray part dark-gray part

Elements HPD1150-R06A8 HPD1150-R06A10 HPD1150-R06B1 HPD1150-R06B2 HPD1151-R05 HPD1153-R03A HPD1153-R03B

Fe (%) 1.26 0.21 1.82 3.18 3.31 2.38 1.48Mn 0.12 0.03 <0.01 <0.01 0.02 <0.01 <0.01Si 10.8 41.6 3.82 5.38 3.16 6.97 10.6Al 0.058 0.053 0.016 0.032 0.058 0.032 0.090Ca 0.08 0.07 0.06 0.06 0.06 0.04 0.04Mg 0.06 0.10 <0.01 <0.01 0.01 <0.01 0.01Na 0.24 0.52 0.03 0.04 0.09 0.08 0.08K 0.02 0.07 <0.01 <0.01 <0.01 <0.01 <0.01Ti <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01P 0.03 <0.01 0.01 0.01 0.03 0.02 0.02Cl 0.33 -- 0.03 0.04 0.12 0.12 0.10S 10.0 1.31 15.1 16.0 21.4 16.3 17.8Ba 37.8 3.61 40.4 34.8 24.9 32.9 19.4(BaSO4) 64.2 6.14 68.7 59.1 42.3 55.9 33.0Pb 0.4 0.1 2.1 1.6 7.2 4.4 4.9(PbS) 0.5 0.1 2.4 1.8 8.3 5.1 5.7Zn 0.4 0.4 2.1 2.1 16.6 8.7 19.1(ZnS) 0.6 0.6 3.1 3.1 24.7 13.0 28.5LOI 1.57 5.74 1.39 3.08 6.79 3.85 7.17Ag (ppm) 21 7 310 399 304 212 314As 220 1430 200 470 370 210 240Bi <0.2 0.8 <0.2 <0.2 0.2 0.6 0.2Br -- -- -- -- -- -- --Cd 1.5 19 30 32 905 491 999Co 1 1 <1 <1 <1 <1 <1Cr <20 <20 <20 <20 <20 <20 <20Cu 330 110 510 510 5580 1630 1830Ga <2 <2 3 3 38 19 45Ge -- -- -- -- -- -- --Hf <2 <2 <2 <2 <2 <2 <2In <0.4 <0.4 <0.4 <0.4 <0.4 <0.4 <0.4Li <20 <20 <20 <20 <20 <20 <20Mo 29 <4 <4 9 59 37 48Nb 3 3 3 3 4 3 3Ni <10 <10 <10 <10 <10 <10 <10Rb <0.4 <0.4 <0.4 <0.4 <0.4 <0.4 <0.4Sb 81 256 340 299 1040 524 765Sn <2 <2 <2 <2 <2 <2 <2Sr 6,930 852 8,460 7,980 5,980 6,180 4,730Th <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2Tl <1 163 29 50 23 40 21U 1.4 0.4 0.1 0.5 0.1 0.2 0.3V 10 <10 <10 <10 <10 <10 <10Zr 16 3 2 5 1 8 <1La (ppm) 12.6 0.5 14.1 11.9 10 10.2 6.9Ce 5.3 <0.2 5.8 5 4.3 3.6 2.8Pr 0.3 <0.1 0.3 0.2 0.2 0.1 0.1Nd 2.3 0.2 2.2 2.1 1.4 1.7 1Sm 1.3 <0.2 1.4 1.2 0.8 1.2 0.6Eu 11.7 1.2 9.9 10.1 6.7 12.1 6.6Gd 0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1Tb <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1Dy <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1Y 2 <1 2 2 1 2 <1Er <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1Yb <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2Lu 0.2 <0.1 0.4 <0.1 0.3 0.2 <0.1Au (ppb) 38 -- 1,850 1,350 14,000 957 3,440Hg 746 14,700 7,580 12,400 28,600 15,500 26,000

2206 HEIN ET AL.

Appendix 2. (Cont.)

Field of large Field of large Field of large Layer from Chimney top Chimney top Layer from Layer from blocks and thick blocks and thick blocks and thick mound with from cluster of from cluster of mound: porous mound: layers, piece of layers, lens from layers, host for dead chimney small chimneys, small chimneys, Type pale-gray part white rim one block another block sample lens in R05A on top pale-gray part dark-gray part

HPD1153- HPD1153- HPD1153- HPD1153- HPD1153- HPD1153- HPD1153- HPD1153-Elements R03C R03F R04 R05A R05B R06 R07A R07B

Fe (%) 1.92 2.43 18.9 19.1 0.68 6.80 3.92 2.81Mn <0.01 0.04 0.01 <0.01 <0.01 0.02 <0.01 <0.01Si 9.68 3.34 -- 1.35 6.03 4.20 1.65 5.10Al 0.037 0.021 0.33 0.106 0.074 0.074 0.016 0.037Ca 0.06 0.12 <0.2 0.01 0.06 0.04 0.07 0.06Mg 0.01 0.07 0.02 <0.01 0.01 0.01 <0.01 <0.01Na 0.06 0.22 -- 0.02 0.04 0.08 0.02 0.07K <0.01 <0.01 <0.2 0.01 <0.01 0.02 <0.01 0.02Ti <0.01 <0.01 <0.02 <0.01 <0.01 <0.01 <0.01 <0.01P 0.01 0.06 <0.02 0.25 0.01 0.03 0.04 0.03Cl 0.08 0.32 -- 0.22 0.03 0.09 0.03 0.07S 14.6 13.2 26.1 31.8 12.9 28.3 19.9 21.1Ba 33.4 36.6 3.44 2.42 42.0 10.2 22.8 16.6(BaSO4) 56.8 62.2 5.80 4.11 71.4 17.3 38.7 28.2Pb 1.9 1.8 0.2 <0.1 <0.1 10.2 16.2 12.9(PbS) 2.2 2.1 0.23 NA NA 11.8 18.7 14.9Zn 5.3 1.4 0.9 0.1 0.6 23.1 11.4 19.0(ZnS) 7.9 2.1 1.4 0.1 0.9 34.4 17.0 28.3LOI 3.23 2.07 -- 16.0 1.2 11.7 5.03 6.73Ag (ppm) 182 46 83 92 41 358 483 441As 160 260 2675 1700 240 1060 1110 460Bi <0.2 <0.2 1.1 1.7 <0.2 <0.2 0.2 <0.2Br -- -- -- -- -- -- -- --Cd 223 3 81 9 20 1120 726 1150Co <1 <1 163 10 <1 <1 <1 <1Cr <20 <20 <20 <20 <20 <20 <20 <20Cu 880 940 47800 282000 5220 8150 12000 7280Ga 6 6 13 <2 3 29 14 28Ge -- -- 4 -- -- -- -- --Hf <2 <2 <2 <2 <2 <2 <2 <2In <0.4 <0.4 2.7 48 0.5 <0.4 <0.4 <0.4Li <20 <20 <20 <20 <20 <20 <20 <20Mo 25 25 150 7 4 65 14 90Nb 3 3 <2 3 3 3 3 3Ni <10 <10 <10 <10 <10 <10 <10 <10Rb <0.4 <0.4 0.9 <0.4 <0.4 <0.4 <0.4 <0.4Sb 305 75 96 71 104 1,320 986 1,070Sn <2 <2 <2 2 <2 <2 2 <2Sr 6,510 8,820 812 469 11,500 2,670 7,380 4,940Th <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2Tl 59 33 160 43 13 58 10 27U 0.3 0.9 3.7 0.3 0.6 0.1 <0.1 0.3V <10 <10 30 <10 <10 <10 <10 <10Zr 1 4 2 6 1 <1 1 2La (ppm) 10.3 13.6 2.9 0.9 14 2.4 16.2 4.3Ce 3.9 6 3.3 0.8 5.8 0.5 8.1 1.6Pr 0.1 0.3 0.3 <0.1 0.2 <0.1 0.4 <0.1Nd 1.8 2.3 1.1 0.7 2.4 0.4 1.8 0.7Sm 1.2 1.2 0.3 <0.2 1.6 0.2 0.8 0.5Eu 11.2 14.5 0.1 0.7 17.9 2.9 8.5 5.6Gd <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1Tb <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1Dy <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1Y 2 2 <1 <1 2 <1 <1 <1Er <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1Yb <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2 <0.2Lu 0.3 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1Au (ppb) 771 639 I.S. 2,600 899 19,000 7,010 13,000Hg 10,500 11,800 8,420 13,800 12,500 54,600 54,900 37,500

1 Dash means not analyzed2 Compounds in parenthesis are calculated using the above metal content3 The following elements are at or below detection limits (ppm in parenthesis): Be (0.1), Cs (0.2), Sc (0.01), Se (1), Ta (0.5), Te (0.05), W (1), Ho (0.05),

Tm (0.05)