@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Science, Faculty of"@en, "Earth and Ocean Sciences, Department of"@en ; edm:dataProvider "DSpace"@en ; ns0:identifierCitation "Barker, A. K.; Coogan, Lawrence A.; Gillis, Kathryn M.; Weis, Dominique. (2008). Strontium isotope constraints on fluid flow in the sheeted dike complex of fast spreading crust: pervasive fluid flow at Pito Deep. Geochemistry, Geophysics, Geosystems, 9(6), Q06010"@en ; dcterms:contributor "University of British Columbia. Pacific Centre for Isotopic and Geochemical Research"@en ; ns0:rightsCopyright "Weis, Dominique"@en ; dcterms:creator "Weis, Dominique"@en ; dcterms:issued "2016-12-01T11:44:23"@en, "2008"@en ; dcterms:description """Fluid flow through the axial hydrothermal system at fast spreading ridges is investigated using the Sr-isotopic composition of upper crustal samples recovered from a tectonic window at Pito Deep (NE Easter microplate). Samples from the sheeted dike complex collected away from macroscopic evidence of channelized fluid flow, such as faults and centimeter-scale hydrothermal veins, show a range of 87Sr/86Sr from 0.7025 to 0.7030 averaging 0.70276 relative to a protolith with 87Sr/86Sr of ∼0.7024. There is no systematic variation in 87Sr/86Sr with depth in the sheeted dike complex. Comparison of these new data with the two other localities that similar data sets exist for (ODP Hole 504B and the Hess Deep tectonic window) reveals that the extent of Sr-isotope exchange is similar in all of these locations. Models that assume that fluid-rock reaction occurs during one-dimensional (recharge) flow lead to significant decreases in the predicted extent of isotopic modification of the rock with depth in the crust. These model results show systematic misfits when compared with the data that can only be avoided if the fluid flow is assumed to be focused in isolated channels with very slow fluid-rock exchange. In this scenario the fluid at the base of the crust is little modified in 87Sr/86Sr from seawater and thus unlike vent fluids. Additionally, this model predicts that some rocks should show no change from the fresh-rock 87Sr/86Sr, but this is not observed. Alternatively, models in which fluid-rock reaction occurs during upflow (discharge) as well as downflow, or in which fluids are recirculated within the hydrothermal system, can reproduce the observed lack of variation in 87Sr/86Sr with depth in the crust. Minimum time-integrated fluid fluxes, calculated from mass balance, are between 1.5 and 2.6 × 106 kg m−2 for all areas studied to date. However, new evidence from both the rocks and a compilation of vent fluid compositions demonstrates that some Sr is leached from the crust. Because this leaching lowers the fluid 87Sr/86Sr without changing the rock 87Sr/86Sr, these mass balance models must underestimate the time-integrated fluid flux. Additionally, these values do not account for fluid flow that is channelized within the crust. An edited version of this paper was published by AGU. Copyright 2008 American Geophysical Union."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/39090?expand=metadata"@en ; skos:note "Strontium isotope constraints on fluid flow in the sheeted dike complex of fast spreading crust: Pervasive fluid flow at Pito Deep A. K. Barker, L. A. Coogan, and K. M. Gillis School of Earth and Ocean Sciences, University of Victoria, P.O. Box 3055 STN CSC, Victoria, British Columbia V8W 3P6, Canada (akbarker@uvic.ca) D. Weis Pacific Centre for Isotopic and Geochemical Research, Department of Earth and Ocean Sciences, University of British Columbia, 6339 Stores Road, Vancouver, British Columbia V6T 1Z4, Canada [1] Fluid flow through the axial hydrothermal system at fast spreading ridges is investigated using the Sr- isotopic composition of upper crustal samples recovered from a tectonic window at Pito Deep (NE Easter microplate). Samples from the sheeted dike complex collected away from macroscopic evidence of channelized fluid flow, such as faults and centimeter-scale hydrothermal veins, show a range of 87Sr/86Sr from 0.7025 to 0.7030 averaging 0.70276 relative to a protolith with 87Sr/86Sr of 0.7024. There is no systematic variation in 87Sr/86Sr with depth in the sheeted dike complex. Comparison of these new data with the two other localities that similar data sets exist for (ODP Hole 504B and the Hess Deep tectonic window) reveals that the extent of Sr-isotope exchange is similar in all of these locations. Models that assume that fluid-rock reaction occurs during one-dimensional (recharge) flow lead to significant decreases in the predicted extent of isotopic modification of the rock with depth in the crust. These model results show systematic misfits when compared with the data that can only be avoided if the fluid flow is assumed to be focused in isolated channels with very slow fluid-rock exchange. In this scenario the fluid at the base of the crust is little modified in 87Sr/86Sr from seawater and thus unlike vent fluids. Additionally, this model predicts that some rocks should show no change from the fresh-rock 87Sr/86Sr, but this is not observed. Alternatively, models in which fluid-rock reaction occurs during upflow (discharge) as well as downflow, or in which fluids are recirculated within the hydrothermal system, can reproduce the observed lack of variation in 87Sr/86Sr with depth in the crust. Minimum time-integrated fluid fluxes, calculated from mass balance, are between 1.5 and 2.6  106 kg m2 for all areas studied to date. However, new evidence from both the rocks and a compilation of vent fluid compositions demonstrates that some Sr is leached from the crust. Because this leaching lowers the fluid 87Sr/86Sr without changing the rock 87Sr/86Sr, these mass balance models must underestimate the time-integrated fluid flux. Additionally, these values do not account for fluid flow that is channelized within the crust. Components: 11,828 words, 10 figures, 2 tables. Keywords: Sr isotopes; sheeted dike complex; hydrothermal alteration; ocean crust. Index Terms: 3017 Marine Geology and Geophysics: Hydrothermal systems (0450, 1034, 3616, 4832, 8135, 8424); 3035 Marine Geology and Geophysics: Midocean ridge processes; 3616 Mineralogy and Petrology: Hydrothermal systems (0450, 1034, 3017, 4832, 8135, 8424). G3GeochemistryGeophysicsGeosystems Published by AGU and the Geochemical Society AN ELECTRONIC JOURNAL OF THE EARTH SCIENCES Article Volume 9, Number 6 24 June 2008 Q06010, doi:10.1029/2007GC001901 ISSN: 1525-2027 Copyright 2008 by the American Geophysical Union 1 of 19 Received 15 November 2007; Revised 14 April 2008; Accepted 24 April 2008; Published 24 June 2008. Barker, A. K., L. A. Coogan, K. M. Gillis, and D. Weis (2008), Strontium isotope constraints on fluid flow in the sheeted dike complex of fast spreading crust: Pervasive fluid flow at Pito Deep, Geochem. Geophys. Geosyst., 9, Q06010, doi:10.1029/ 2007GC001901. 1. Introduction [2] Hydrothermal circulation at mid-ocean ridges plays a critical role in controlling the composition of the world’s oceans, provides the energy and nutrients for extremophile organisms and generates ore deposits that can be of economic importance. Understanding the hydrothermal modification of the bulk composition of the crust is also important for quantifying the chemical inputs into subduc- tion zones and recycling within the mantle [e.g., Hofmann and White, 1982]. However, our under- standing of the subsurface processes operating in ridge axis hydrothermal systems is limited. [3] At intermediate to fast spreading ridges, where the crustal structure is relatively simple, cold seawater migrates downward through the lava pile and sheeted dike complex and is heated and chemically modified through mineral precipitation and reaction with the crust as it goes. Eventually the chemically modified hot hydrothermal fluid discharges upward and exits the crust, either at focused vents or more diffusely after mixing with cooler pore fluids. It is widely believed that the main modification of the mineralogy and compo- sition of the sheeted dike complex occurs within the recharge portion of the system [e.g., Alt, 1995; Teagle et al., 2003] with discharge being focused into small areas. The total fluid mass that passes through these systems, and the heat and chemical fluxes carried by the fluid, are poorly constrained. Additionally the extent of variability in these parameters is unclear. For example does the fluid flux depend on spreading rate or does it vary temporally at any given location? One way to better understand these systems is through the study of samples of the oceanic crust. [4] Much previous study has concentrated on the record of hydrothermal circulation in rocks recov- ered by deep drilling, especially the core recovered from ODP Hole 504B [e.g., Alt et al., 1996b; Laverne et al., 2001; Teagle et al., 1998a, 1998b, 2003; Bach et al., 2003]. However, drill cores provide a largely one-dimensional view of the axial hydrothermal system and reveal little about the spatial variability of these systems. Studies of a tectonic exposure of the upper ocean crust at Hess Deep [Gillis, 1995; Gillis et al., 2001], and of on land exposures of oceanic crust formed at abnormal spreading centers (ophiolites [e.g., Schiffman et al., 1987]), have demonstrated the importance of three- dimensional variability in hydrothermal systems. The approach used here to understand the processes operating in axial hydrothermal systems is to study a ‘‘tectonic window’’ at Pito Deep, in the southern Pacific. The crust has been dissected by faulting forming a ‘‘tectonic window’’ where laterally con- tinuous upper crustal exposures allow the spatial and temporal variability in fluid flow to be inves- tigated. Additionally, comparison of the hydrother- mal fluxes through the sheeted dike complex at Pito Deep with those calculated from studies of previous crustal sections (ODP Hole 504B, Hess Deep) allows the global variability in hydrothermal fluxes to be investigated. 1.1. Sr Isotope Exchange [5] The strontium isotopic composition of the oceanic crust currently provides arguably the most quantitative information about the amount of hy- drothermal fluid that the crust has reacted with. This is in part due to the contrast in 87Sr/86Sr between fresh ocean crust (Sr  100 ppm; 87Sr/86Sr  0.7024–0.7025; PETDB: http://www.petdb.org) and seawater (Sr  8 ppm, 87Sr/86Sr  0.7091 [Hess et al., 1986]). Additionally, the mobility of Sr in the presence of fluid at greenschist facies conditions promotes isotopic exchange between the hydrothermal fluid and rock it encounters. As discussed in more detail later, this means that the modification of the 87Sr/86Sr of the ocean crust allows the extent and variability of interaction between the fluid and crust to be assessed [e.g., Bickle and Teagle, 1992; Bickle et al., 1998; Davis et al., 2003; Bach et al., 2003; Teagle et al., 2003; Gillis et al., 2005]. Likewise, the 87Sr/86Sr of hydrothermal fluids venting from the crust provide insight into the amount of rock that the fluid has reacted with. At intermediate to fast spreading ridges the mean 87Sr/86Sr of vent fluids is 0.7038 [Bach and Humphris, 1999; Von Damm, 1990, Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 2 of 19 1995; Palmer, 1992; Palmer and Edmond, 1989; Merlivat et al., 1987; Hinkley and Tatsumoto, 1987; Piepgras and Wasserburg, 1985; Albarède et al., 1981]. These fluid isotopic compositions can be used to calculate the amount of rock that the fluid reacted with but, as shown by Berndt et al. [1988], these calculations suffer considerable un- certainty in the absence of constraints from the composition of the rocks. [6] Here we use the 87Sr/86Sr of a suite of samples recovered on a recent cruise to Pito Deep, together with supporting petrography and major and trace element analyses, to investigate the scale of vari- ability in the extent of pervasive hydrothermal fluid flow at intermediate to fast spreading ridge axes. We focus on the alteration of the sheeted dike complex because this is the unit in which there is evidence for high-temperature alteration at condi- tions similar to those required to produce fluids with the compositions observed at hydrothermal vents [e.g., Seyfried, 1987; Alt, 1995; Gillis et al., 2001]. We show that the extent of Sr-isotopic exchange between the crust and hydrothermal fluids is very similar in all areas studied at inter- mediate to fast spreading ridges and requires a minimum of 1.5 to 2.6  106 kg m2 of fluid to circulate through the crust. Additionally we sug- gest that the lack of variation in 87Sr/86Sr with depth in the crust is inconsistent with exchange of Sr between the fluid and rock occurring only during recharge and instead that reaction during discharge is required by the data. 1.2. Defining Pervasive Fluid Flow [7] In this study we focus on ‘‘pervasive’’ fluid flow through the sheeted dike complex. We use this term to represent the fluid flow that leads to alteration that is relatively homogeneously distrib- uted within the sheeted dike complex and contrasts with the alteration associated with focused fluid flow. The latter is generally associated with fluid flow in fault zones and mineralized veins that are tens of millimeters wide and is the subject of an ongoing study (A. K. Barker et al., manuscript in preparation, 2008). We consider this to be a logical division because this probably reflects two differ- ent permeability systems within the axial hydro- thermal system, although, there must be a broad continuum between the two. We use the somewhat arbitrary definition that samples are associated with focused fluid flow if dive footage shows them to be within 2 m of a fault, breccia zone or regions of extensive hydrothermal veining at the sampling location on the seafloor. Additionally, samples containing veins 5 mm wide are also considered to have been associated with focused fluid flow. All other samples are assumed to have been exposed only to pervasive fluid flow. On the basis of these criteria 53 samples were chosen for this study to represent the pervasive hydrothermal fluid flow at Pito Deep. [8] An important aspect of this study is the com- parison of the samples recovered from Pito Deep with those from other areas, hence it is important to note the differences in sampling strategy in the different study areas. The most comparable sample suite is that from Hess Deep [Gillis et al., 2005]. Although these authors did not stringently filter their samples to avoid any from near fault zones they did not analyze samples from directly within fault zones. In contrast, samples from the full range of extent of deformation and alteration from ODP Hole 504B have been analyzed [Alt et al., 1996b]. Further comparison between these different study areas is given in Table 1. 2. Geological Background [9] Pito Deep is located at the northeast corner of the Easter microplate (Figure 1a) [Francheteau et al., 1988], where a tectonic window exposes 3 Ma old ocean crust formed at a fast spreading ridge at a full spreading rate of 142 mm/a [Hey et al., 1995]. Rifting and rotation at the north edge of the Easter microplate faulted and exposed the ocean crust at Pito Deep within the last million years [Hey et al., 2002]. During cruise AT11-23 of the R/V Atlantis in February 2005 to Pito Deep, two escarpments were sampled by the ROV Jason II and the Alvin submersible by prizing samples from outcrop leading to in situ sampling of multiple densely sampled dive transects (Figure 1). These two escarpments are referred to as Areas A and B; the former is located west of the latter. Volcanic rocks make up the upper 300–500 m of each crustal section and are underlain by 650–1100 m of sheeted dikes [Heft et al., 2008; Karson and Pito Deep Shipboard Scientific Party, 2005]. The base of the sheeted dike complex is only exposed in Area B where 900 m of gabbros are exposed [Perk et al., 2007]. The escarpments trend perpen- dicular to the strike of the dikes providing a cross- section through the crust [Hey et al., 2002; Karson and Pito Deep Shipboard Scientific Party, 2005]. Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 3 of 19 [10] Samples for this study were selected to pro- vide as complete a vertical profile through the upper ocean crust as possible. In Area A, three adjacent and overlapping dives provided lateral coverage of 500 m and a depth range of 1000 m; the sample density in the lower 200 m of dives J2-119-1 and J2-119-2 was restricted by the occurrence of faulting and macroscopic veining (Figure 1a). In Area B, five adjacent and over- lapping dives provided lateral coverage of 1000 m at near-constant depth between 100 and 400 m below the lava-dike transition and a depth range of 1300 m. There is a gap in sampling between 450 and 1050 m in Area B J2-123-4 due to talus concealing the exposure (Figure 1b) [Karson and Pito Deep Shipboard Scientific Party, 2005]. An important datum for this study is the lava-dike transition, which is defined here as the depth of the deepest lava in each dive transect and data are plotted with depths relative to this transi- tion in all figures. 3. Analytical Methods [11] Weathered edges and alteration haloes were removed from the samples prior to jaw crushing; subsequently samples were ground in an agate mill. Trace element analyses were performed on whole rock powders, 100 mg of sample was digested with HF-HNO3 and 8N HNO3, and diluted by a factor of 1:1000 in 2% HNO3 solution. Trace elements were analyzed using a Thermo X Series II induc- tively coupled plasma mass spectrometry (ICP- MS) following the approach of Eggins et al. [1997] at the University of Victoria. BIR-1 was routinely measured to monitor for drift, and DNC-1, W2-A, BCR-2, BHVO-2 rock standards were analyzed for calibration at the start and end of analytical sessions (auxiliary material1 Data Set S1). [12] The 87Sr/86Sr of lavas and dikes were mea- sured on unleached whole rock powders on 40 of the 53 samples. The Sr isotope ratios of epidote separates from two quartz-epidote veins were also determined. An estimate of the 87Sr/86Sr of fresh ocean crust at Pito Deep and Hess Deep was made on the basis of analysis of leached plagio- clase separates of the least altered gabbros from the underlying crust [Perk et al., 2007; Coogan et al., 2002; Francheteau et al., 1990]. Plagioclase separates were leached in hot 6N HCl for 30 min prior to digestion. Strontium was separated by standard cation exchange techniques [Weis et al., 2006] and measured using a Triton thermal ion- ization mass spectrometry at the Pacific Centre for Isotopic and Geochemical Research, University of British Columbia. The NBS 987 standard gave 87Sr/86Sr of 0.710251 ± 0.000017 (2 s.d., n = 26), the maximum difference between duplicates was ±0.000022 with an average reproducibility Table 1. Summary of Characteristics of Sheeted Dikes at Pito Deep, Hess Deep, and ODP Hole 504Ba Pito Deep Hess Deep 504B Pito Deep Area A Area B Age (Ma) 3 1 6.9 Spreading rate (mm a1) 142 130 66 Thickness of Sheeted Dikes (m) 650–1100 200–1000 >1050 Protolith Mg# (mean) 55 50 - 60 55 Mg# range 42–68 52–66 63–74 42–68 49–60 Anb <0.7 <0.7 >0.7 Extent of cpx alteration % <10–>85 <10–40 5 – 40 Mean Sr (ppm) 95 90 60 100 95 87Sr/86Sr 0.70239–0.70242 0.70243–0.70245 0.70245a a The Sr isotopic composition of fresh rock proximal to ODP Hole 504B from a latitude of 1.75N and longitude of 85.17W has a mean 87Sr/86Sr of 0.70245 (n = 3 [Verma and Schilling, 1982; Schilling et al., 2003]). Sources: This study, Heft et al. [2008], Pollock et al. (submitted manuscript, 2008), Karson et al. [2002], Karson and Pito Deep Shipboard Scientific Party, 2005; Teagle et al. [2003], Bach et al. [2003], Gillis et al. [2001, 2005], Laverne et al. [2001], Alt et al. [1996a, 1996b], Hékinian et al. [1996], Hilgen et al. [1995], Lonsdale [1988], Martinez et al. [1991], Hey et al. [1977]. b Anorthite content of plagioclase feldspar (An = Ca/[Ca + Na] molar). 1Auxiliary materials are available at ftp://ftp.agu.org/apend/gc/ 2007gc001901. Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 4 of 19 ±0.000012 (±8, 1 s.d.), the Sr blank measured over the course of the study was 13 pg. 4. Results 4.1. Petrography of the Lavas and Sheeted Dike Complex [13] Volcanic rocks from Pito Deep have visually estimated abundances of secondary minerals in thin section ranging from 10 to 40% of the rock. Clay minerals and oxyhydroxide replace the primary minerals in the upper volcanics and amphibole and albite in the lava-dike transition zone. Sheeted dikes at Pito Deep range from aphyric basalts to porphyritic and equigranular basalts with minimal interstitial phases. The samples range in grain size from very fine to 0.4 mm, perhaps representing the center of sheeted dikes opposed to chilled margins. Primary plagioclase is altered to second- ary plagioclase ± amphibole, ±chlorite, ±Fe-oxides and rare epidote. Clinopyroxene ranges from fresh through partially replaced by amphibole, ±chlorite, Figure 1. (a) Locations of dive transects and samples studied in Area A. Inset shows map of Pito Deep at northeast of Easter microplate and relative locations of Hess Deep and ODP Hole 504B. (b) Locations of dive transects and samples studied in Area B [Karson and Pito Deep Shipboard Scientific Party, 2005]. Lava-dike transition is gradual with dikes intruding lavas to different extents (Pollock et al., submitted manuscript, 2008). Faults occur at high angles and subparallel to the strike of the sheeted dikes, with examples of sheeted dikes crosscutting fault zones indicating formation at the ridge axis [Karson and Pito Deep Shipboard Scientific Party, 2005]. Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 5 of 19 ±minor Fe-oxides and rare secondary clinopyrox- ene. Amphibole and chlorite often occur together in the groundmass. In general the alteration assemblages are dominated by amphibole (auxilia- ry material Data Set S1) [Heft et al., 2008]. Similar alteration to amphibole-dominated assemblages is observed in the sheeted dikes from ODP Hole 504B with chlorite dominated assemblages con- centrating in the upper 400 m of the sheeted dike complex [Laverne et al., 1995, 2001; Vanko et al., 1996; Alt et al., 1989, 1996b], whereas sheeted dikes from Hess Deep show more variability in alteration assemblages with a greater occurrence of chlorite dominated (57%) than amphibole domi- nated assemblages (43%, n = 26) in samples with Sr isotope data [Gillis et al., 2005]. [14] Samples exhibit patches of contrasting alter- ation, which tend toward chlorite-dominated assemblages. Patches are typically 1–2 mm in diameter and are composed of chlorite, ±plagio- clase, ±quartz, ±sulphides, ±epidote, ±amphibole. A patch bearing sample typically hosts a single patch. Patches are more prevalent in Area A with 20% of samples hosting patches as opposed to 15% in Area B (n = 20 and 33, respectively). Veins are common in thin section with 50% of the samples hosting from one to ten veins of 0.015mm to 0.7mm width. Vein mineralogy includes amphibole, chlo- rite, quartz, epidote and sulphides. Epidote most commonly occurs in association with patches or veins but also occurs as an alteration product of plagioclase away from patches and veins. Thin veins 1 mm wide visible in hand specimen are more abundant in Area B, which hosts veins in 58% of samples whereas Area A hosts veins in 45% of samples (n = 33 and 20, respectively). [15] The extent of alteration in samples from Pito Deep was visually estimated from the modal pro- portion of secondary minerals. There is a wide petrographic range in the extent of total alteration at Pito Deep from <10 to >85% [Heft et al., 2008; this study]. Areas A and B show similar extents of total alteration averaging 34% (±17, 1 s.d.). Clino- pyroxene is similarly altered between <10 and >95%, with an average of 33% (±22, 1 s.d.). Heft et al. [2008] show that there are no systematic spatial trends in alteration assemblages and that degree of alteration and amphibole abundance do not correlate with depth. 4.2. Protolith Composition [16] The dikes and lavas at the Pito Deep are typical moderately fractionated N-MORB [Hékinian et al., 1996; Heft et al., 2008; this study; M. A. Pollock et al., Compositions of dikes and lavas from the Pito Deep Rift: Implications for crustal accretion at superfast spreading centers, submitted to Journal of Geophysical Research, 2008]. There are small differences in the primary (magmatic) composition of the dikes between Areas A and B; the mean Mg# (Mg# = Mg/(Mg + Fe) molar) is 60 ± 13 (1 s.d.) in Area A and 55 ± 6 (1 s.d.) in Area B [Hékinian et al., 1996; Heft et al., 2008; Pollock et al., submitted manuscript, 2008]. The Sr contents of the sheeted dike samples in Area A average 100 ± 10 (1 s.d.) ppm, whereas Sr concentrations in samples at Area B are slightly lower with an average of 95 ± 6 (1 s.d.) ppm Sr. These small differences in protolith composition are expected to be insignificant in controlling the exchange of Sr-isotopes between the crust and hydrothermal fluids. 4.3. Strontium Isotopic Composition of the Ocean Crust From Pito Deep [17] The extent of hydrothermal enrichment of 87Sr/86Sr in the Pito Deep ocean crust requires knowledge of the 87Sr/86Sr of the local fresh protolith. We have determined the protolith 87Sr/86Sr from leached plagioclase separated from the least altered gabbros from Pito Deep and 87Sr/86Sr of proximal MORB currently at the East Pacific Rise (EPR) between 22 and 23S. The plagioclase separates have 87Sr/86Sr of 0.70237 to 0.70239 (auxiliary material Data Set S1) and local fresh MORB have 87Sr/86Sr between 0.70239 and 0.70242 [Mahoney et al., 1994]. On the basis of these data we assume a range of initial (i.e., mag- matic) 87Sr/86Sr of 0.70237 to 0.70242 for Pito Deep ocean crust. To supplement the data set of Gillis et al. [2005] on altered dikes and lavas fromHess Deep we have analyzed plagioclase separates from gabbros to define the composition of fresh crust at Hess Deep. Plagioclase separates from least altered gabbros at Hess Deep have 87Sr/86Sr of 0.70243 and 0.70245 (auxiliary material Data Set S1). [18] The Sr isotopic composition of six Pito Deep lavas sampled from a 280 m depth interval range from 87Sr/86Sr of 0.70257 to 0.70363 (Figure 2). The Sr isotope compositions of the Pito Deep lavas do not correlate with alteration assemblages, extent of total or clinopyroxene alteration or alkali ele- ment concentration. For comparison two lavas with Sr isotope data from younger crust at Hess Deep have 87Sr/86Sr of 0.70255 and 0.70260 [Gillis et al., 2005], similar to the lowest 87Sr/86Sr values Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 6 of 19 observed at Pito Deep. The extensive analyses for the older lavas at ODP Hole 504B, sampled over a 570 m depth interval range in 87Sr/86Sr from 0.70255 to 0.70488 [Bach et al., 2003; Teagle et al., 1998a, 1998b, 2003; Alt et al., 1996a, 1996b]. Wide ranges in Sr isotopic composition as ob- served are expected of the variable low temperature alteration from heterogeneous fluid flow that occurs in high-permeability volcanics [Alt et al., 1996b; Kusakabe et al., 1989]. [19] The Sr isotopic composition of the sheeted dikes at Pito Deep range from 87Sr/86Sr of 0.70252 to 0.70295 (auxiliary material Data Set S1 and Figure 2). There is no observed change in 87Sr/86Sr with depth and the deepest dike samples at 1100 m below the lava-dike transition have 87Sr/86Sr of 0.70265, well above the lowest 87Sr/86Sr and closer to the mean 87Sr/86Sr of 0.70276 for dikes at Pito Deep. The sheeted dikes in the two escarpments sampled at Pito Deep have similar Sr isotopic compositions; in Area A sheeted dikes have 87Sr/86Sr of 0.70252 to 0.70293 and in Area B the range is from 0.70265 to 0.70295 (Figure 2). All of the sheeted dikes at Pito Deep have Sr isotope compositions more radiogenic than that of the pro- tolith (87Sr/86Sr between 0.70237 to 0.70242) dem- onstrating that all samples have exchanged Sr with hydrothermal fluids. There are no spatial groupings of Sr isotope composition instead the entire range of 87Sr/86Sr in the sheeted dike complex can be found in adjacent samples and transects. [20] Epidote was separated from two centimeter- scale quartz-epidote veins from Area B that occur 220 m and 370 m below the lava-dike transition. These have 87Sr/86Sr of 0.70382 and 0.70413 similar to epidote veins from ODP Hole 504B (87Sr/86Sr of 0.7034 to 0.7038 [Teagle et al., 1998b]). Proximal sheeted dikes have 87Sr/86Sr of 0.70270 and 0.70271 (Samples 4086-1711 at 2 m and 022205-0941 at 15 m from vein). These quartz-epidote veins have 87Sr/86Sr similar to av- erage vent fluids at intermediate to fast spreading ridges suggesting that these epidote veins were precipitated after the fluid has undergone most or all of its Sr-isotopic exchange with the crust. 4.4. Mineralogical Controls on the Sr- Isotopic Composition of the Sheeted Dikes [21] If the exchange of Sr between a hydrothermal fluid and the crust is dominantly controlled by the minerals forming in the rock then this would have to be accounted for in modeling the Sr-isotopic composition of the crust to better understand hy- drothermal circulation. For example, Gillis et al. [2005] found that, compared to the protolith, sam- ples from Hess Deep that contained epidote were elevated in Sr and in 87Sr/86Sr whereas those that did not contain epidote were depleted in Sr and enriched somewhat less in 87Sr/86Sr. The enrich- ment of Sr in rocks containing epidote can be explained by the high partition coefficient for Sr into epidote assuming that it grew in the presence of sufficient fluid to supply Sr. Seven of the sheeted dikes from Pito Deep studied here contain epidote. These are not significantly enriched in either Sr concentration or 87Sr/86Sr relative to the other samples (Figure 2). There is also no correla- tion of the modal proportions of chlorite or amphi- bole, or grain size, with 87Sr/86Sr. Likewise, the alkali elements that are commonly enriched in Figure 2. Values of 87Sr/86Sr with depth below the lava-dike transition for lavas and dikes at Pito Deep. Area A and B samples are separated into open and filled symbols, respectively. Inset shows mean 87Sr/86Sr ± 1 s.d. with depth at 100 m depth intervals for sheeted dikes from Areas A and B. Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 7 of 19 minerals that form at low temperatures, such as clays, show no correlation with 87Sr/86Sr. It is difficult in thin section to accurately assess the amount of secondary (albitic) plagioclase in a sample. Thus to assess the role of albitization in Sr-mobility we use bulk-rock chemical proxies for this process. There are no correlations between Na, Ca, or ratios such as Na/Y, Ca/Na and Sr concen- tration, indicating that albitization is not a factor influencing Sr behavior during alteration. 5. Data Analysis and Modeling [22] With the data presented above there are now three areas of oceanic crust formed at intermediate to fast spreading ridges that the Sr-isotopic com- position of the sheeted dike complex has been determined for: Pito Deep, Hess Deep and ODP Hole 504B. In this section we use these three data sets to address three questions: (1) How variable is the extent of Sr-isotopic exchange between hydro- thermal fluids and the crust in these areas? (2) What range of time-integrated fluid fluxes through the sheeted dike complex do these data suggest? (3) Are these data consistent with fluid-rock reaction occurring within the recharge portion of the hydro- thermal system? However, before addressing these issues we consider whether Sr behaves conserva- tively within the axial hydrothermal system, as has previously been proposed, or not. 5.1. How Mobile is Strontium During Hydrothermal Circulation? [23] In order to use Sr isotopes to trace the fluid- rock reaction we need to understand the controls on the behavior of Sr concentration during hydrother- mal alteration. The concentration of Sr in mid-ocean ridge basalts (MORB) remains approx- imately constant during differentiation so that de- fining fractionation trends is not simple. We consider the variations between Sr and REE illus- trated as SrN/NdN and EuN/EuN*, where concen- trations are normalized to chondrite (N) and EuN/EuN* is the Europium anomaly which repre- sents deviations in EuN with respect to neighboring REEs, SmN and GdN. A primitive magma has SrN/ NdN and EuN/EuN* of 1 and crystallization of plagioclase fractionates Sr from Nd, and Eu2+ from Sm and Gd causing the Sr and Eu content of the melt to decrease relative to the other REE. Figure 3 illustrates the curved trend expected from magmatic differentiation between SrN/NdN and EuN/EuN*. Comparing Pito Deep samples to this magmatic differentiation trend allows assessment of the effects of magmatic differentiation and hydrothermal alter- ation on the Sr concentration of samples (Figure 3). Hydrothermal leaching or precipitation of Sr is expected to result in vertical deviations from this trend due to the immobility of REE during hydro- thermal alteration [e.g., Bau, 1991]. [24] Lavas from Pito Deep generally have SrN/NdN at given EuN/EuN* close to the magmatic trend, with just one showing elevated SrN/NdN over the magmatic trend, indicating gain of Sr. Sheeted dikes from Area A also have SrN/NdN at given EuN/EuN* close to the magmatic trend, with one sample showing significant Sr loss and another Sr gain. Sheeted dike samples from Area B have a wider range in deviation of SrN/NdN from the magmatic trend with 45% of the samples showing Sr depletion (>0.05 units below the magmatic trend which is equivalent to 9 ppm loss of Sr). Alto- gether, strontium is depleted in 30% of the sheeted dikes, depleted samples show no consistent spatial distribution. Strontium depletion does not correlate with 87Sr/86Sr unlike in samples from Hess Deep [Gillis et al., 2005]. However, as with samples from Hess Deep it is clear that Sr is Figure 3. SrN/NdN versus EuN/EuN* for sheeted dikes from Pito Deep, Areas A and B. Black line marks the magmatic differentiation trend; gray shaded area marks deviation of ±0.05 SrN/NdN from the magmatic differentiation trend. Hydrothermal mobilization of Sr is represented by deviations in SrN/NdN (>0.05) at constant EuN/EuN* from the magmatic trend. Magmatic trend is calculated by equilibrium fractional crystal- lization of plagioclase with Kd(plag/liq) for Sr = 1.554, Nd = 0.014, Eu = 0.332, Sm = 0.009, and Gd = 0.007 [McKay et al., 1994]. Variations in SrN/NdN do not correlate with 87Sr/86Sr, indices of albitization such as Na, alteration degree, or assemblages. Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 8 of 19 leached from the sheeted dike complex during hydrothermal circulation. [25] Since there is strong evidence that Sr is leached from the sheeted dike complex during hydrothermal alteration we reconsider the previous interpretation that Sr behaves conservatively with the average Sr content of vent fluids matching that of seawater [Palmer and Edmond, 1989; Davis et al., 2003]. Figure 4 shows a compilation of Sr versus Cl data for vent fluids for the EPR and Juan de Fuca ridges. The majority of the samples follow the linear correlation observed by Palmer and Edmond [1989] where Sr behaves conservatively with Cl. However, samples at low Cl deviate from the linear correlation to higher Sr at a given Cl content forming a curved array. Additionally, sam- ples from the Juan de Fuca ridge demonstrate that there are vents where the Sr concentration at a given Cl concentration is substantially higher than the curved trend defined by the majority of the data [Butterfield et al., 1994; Butterfield and Massoth, 1994]. Thus the mean Sr concentration of vent fluids is 40% higher than seawater consistent with the observation that Sr is leached from the sheeted dike complex during high-temperature al- teration at the ridge axis. 5.2. Regional Characteristics of Alteration 5.2.1. Comparable Alteration of Ocean Crust [26] The Sr-isotopic composition of the sheeted dike complex of modern oceanic crust formed at intermediate to fast spreading ridges shows re- markable similarity in the three areas that have been sampled to date (Figure 5). Sheeted dikes at Pito Deep have Sr isotopic compositions between 0.70252 and 0.70295 and dikes from Hess Deep range between 0.70257 and 0.70326 (Figure 5a) [Gillis et al., 2005]. The majority of sheeted dikes from ODP Hole 504B have a range of 87Sr/86Sr from 0.70257 to 0.70307, 17% of sheeted dikes in the upper 450 m of the sheeted dike complex extend beyond this range to 87Sr/86Sr of 0.70535 but none of the deeper dikes have 87Sr/86Sr > 0.7031 (Figure 5b) [Bach et al., 2003; Teagle et al., 1998a, 1998b, 2003; Alt et al., 1996a, 1996b]. The mean 87Sr/86Sr with depth is similar for Pito Deep, Hess Deep and ODP Hole 504B and there is no statistically significant correlation of 87Sr/86Sr with depth in the sheeted dikes at Pito Deep, Hess Deep or ODP Hole 504B (Figures 5a and 5b). The observed similarity in Sr isotopic profiles through the sheeted dike complex exposed at three loca- tions is important and surprising. This implies that sheeted dikes in young intermediate to fast spread- ing ocean crust at different locations with different initial Sr isotopic compositions (see section 5.2.2) are hydrothermally altered to the same narrow range in 87Sr/86Sr (0.7025 to 0.7033). This indi- cates that differences in spreading rate at interme- diate to fast spreading ridges do not result in significant differences in the extent of 87Sr/86Sr exchange between hydrothermal fluid and ocean crust. 5.2.2. Quantifying the Variability in 87Sr/86Sr in Altered Upper Oceanic Crust [27] To further investigate the striking similarity between the extent of Sr-isotopic exchange be- tween the sheeted dike complex and hydrothermal fluids in these three areas we compare the amount of fluid required to produce the observed Sr-isoto- pic modifications. The composition of the protolith in each of the three areas is slightly different, Figure 4. Sr versus Cl concentrations in vent fluids from the EPR and Juan de Fuca ridge. Gray line marks correlation of Palmer and Edmond [1989]. Solid black lines mark seawater composition. Mean vent fluid plotted with a 2s error ellipse to represent statistical uncertainties; in reality, many other factors influence the geological uncertainty. Juan de Fuca and southern EPR vent fluid data plotted as zero-Mg end-member fluids [Butterfield et al., 1994; Butterfield and Massoth, 1994; Charlou et al., 1996]. Vent fluid data from EPR 9–10N are zero-Mg end-member fluids, a number of which have negative Sr; therefore the measured fluids would have higher Sr concentrations [Von Damm, 2000]. Data sources: Von Damm [1990, 1995, 2000], Palmer and Edmond [1989], Palmer [1992], Butterfield et al. [1994], Butterfield and Massoth [1994], Charlou et al. [1996]. Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 9 of 19 therefore closed-system water-to-rock ratio (by mass) has been calculated for all samples to assess the relative alteration from fresh crust [e.g., Taylor, 1977]. Initial-rock Sr concentrations and Sr-isoto- pic compositions for dikes from all three areas are given in Table 1. We assume an initial fluid with 8 ppm Sr with an isotopic composition slightly modified (decreased by 0.0005) from seawater of the crustal age to account for exchange with basaltic Sr within the lava pile [Teagle et al., 2003]. The final fluid is assumed to have 87Sr/86Sr of 0.7038 to match average vent fluids. We assume here, and throughout, that Sr is not significantly taken up by anhydrite in the recharge zone deplet- ing the fluid in Sr. If this assumption is invalid then the calculated fluid/rock ratios, and time-integrated fluid fluxes, will be underestimates. [28] Fluid/rock ratios calculated assuming no change in the Sr concentration in either fluid or rock for the sheeted dike complex at Pito Deep range from 0.38 to 1.41 with a mean of 0.86 ± 0.25 (1 s.d.; Figure 6a). Sheeted dikes from Hess Deep have fluid/rock ratios of 0.24 to 1.97, with mean of 0.77 ± 0.45 (1 s.d.), while sheeted dikes from ODP Hole 504B have fluid/rock ratios of 0.20 to 4.73 with mean of 0.70 ± 0.65 (1 s.d.). Allowing sufficient Sr to be dissolved out of the rock into the fluid to increase the fluid Sr content to 11 ppm (Figure 5) results in fluid/rock ratios that are higher by an average of 0.1. All of the samples from all localities have fluid/rock ratios  0.20, suggesting that all dikes have exchanged Sr with a hydrother- mal fluid on a sample-scale (i.e., no areas greater than a few centimeters experience no fluid flow). 5.2.3. Local Scale Variations at Pito Deep: Areas A and B [29] The sampling of two escarpments (Areas A and B) [Karson and Pito Deep Shipboard Scientific Party, 2005] at Pito Deep allows comparison of pervasive hydrothermal fluid flow in crustal sec- tions approximately 20 km apart. Calculated fluid/ rock ratios for samples from Area B range from 0.64 to 1.38 with mean of 0.89 ± 0.20 (1 s.d.); samples from Area A are slightly more heteroge- neously altered with fluid/rock ratios ranging from Figure 5. (a) Values of 87Sr/86Sr of sheeted dikes with depth for comparison of Pito Deep with Hess Deep. Inset shows mean 87Sr/86Sr ± 1 s.d. for sheeted dikes at 100 m depth intervals for Pito Deep (circles, red line) and Hess Deep (diamonds, green line). (b) Values of 87Sr/86Sr of sheeted dikes with depth for Pito Deep and ODP Hole 504B. Inset shows 87Sr/86Sr mean ± 1 s.d. for sheeted dikes at 100 m depth intervals for Pito Deep (circles, red line) and ODP Hole 504B (squares, blue line). Data sources: Gillis et al. [2005], Bach et al. [2003], Teagle et al. [1998a, 1998b, 2003], Alt et al., 1996a, 1996b; Bickle and Teagle [1992]. Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 10 of 19 0.38 to 1.41 with a mean of 0.82 ± 0.30 (1 s.d.; Figure 6b). 5.3. Estimating the Time-Integrated Fluid Flux [30] We use three approaches to model the time- integrated fluid fluxes (in kg of fluid per m2 of crust) responsible for pervasive fluid flow in the sheeted dike complex at Pito Deep, Hess Deep and ODP Hole 504B: (1) mass balance of closed system fluid/rock ratios; (2) one-dimensional flu- id-flow (recharge) over a vertical profile with kinetically limited isotopic exchange between fluid and rock; and (3) one-dimensional fluid-flow (re- charge) over a vertical profile with diffusive ex- change of Sr-isotopes between the rock and a fluid confined to cracks. The results suggest that the minimum time-integrated fluid fluxes for all areas lies between 1.5 and 2.6  106 kg m2. 5.3.1. Fluid-Rock Mass Balance [31] Time-integrated fluid fluxes (kg m2) can be calculated from fluid/rock ratios by multiplying by the mass of rock per m2 column of crust. This depends on the thickness of the sheeted dike complex (1100 m at Pito Deep, 700 m at Hess Deep and >1050 m at ODP Hole 504B [Karson et al., 2002; Karson and Pito Deep Shipboard Scientific Party, 2005; Alt et al., 1996a]) and density of basaltic dikes (2.7 kg m3). The average closed system fluid/rock ratios for the sheeted dike complexes at Pito Deep, Hess Deep and ODP Hole 504B are 0.86, 0.77 and 0.70, respectively (Figure 6a). Resulting mean time- integrated fluid fluxes are 2.6  106 kg m2, 1.5 106 kgm2 and 2.0 106 kgm2, respectively (Table 2). Although samples from Hess Deep and Pito Deep have similar mean fluid/rock ratios, the sheeted dike complex is thinner at Hess Deep requiring a smaller fluid flux. If Sr loss from the rock to the fluid is included in the computation of the fluid/rock ratio (section 5.2.2) the time-integrated fluid flux is 0.2 to 0.4 kg m2 higher. 5.3.2. Depth-Constrained Isotopic Exchange During Recharge: A Linear Kinetic Approximation [32] The time-integrated fluid fluxes calculated in the previous section provide useful constraints on the amount of fluid that reacted with the sheeted dike complex. It is generally assumed that the reaction of fluid and rock occurs during recharge of the hydrothermal system [e.g., Teagle et al., Figure 6. Histograms of closed system fluid/rock ratios for (a) Pito Deep, Areas A and B, (b) Hess Deep, and (c) ODP Hole 504B. Geometric mean for Pito Deep Total = 0.82, Pito Deep Area A = 0.77, Area B = 0.86, Hess Deep = 0.66, and ODP Hole 504B = 0.55. Closed system fluid/rock ratios calculated from protolith of 87Sr/86Sr = 0.70239, Sr = 95 ppm at Pito Deep; Area A Sr = 100 ppm, whereas Area B Sr = 95 ppm, with seawater of 8 ppm Sr and 87Sr/86Sr = 0.7086 and final vent fluid of 8 ppm Sr and 87Sr/86Sr = 0.7038. Hess Deep protolith has 87Sr/86Sr = 0.70245 with Sr = 90 ppm [Gillis et al., 2005], with seawater of 8 ppm Sr and 87Sr/86Sr = 0.7086 and final vent fluid of 8 ppm Sr and 87Sr/86Sr = 0.7038. ODP Hole 504B protolith has 87Sr/86Sr = 0.70245 and Sr = 60 ppm [Schilling et al., 2003; Teagle et al., 1998a; Verma and Schilling, 1982], with seawater of 8 ppm Sr and 87Sr/86Sr = 0.7084 [Teagle et al., 2003] and final vent fluid of 8 ppm Sr and 87Sr/86Sr = 0.7038. Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 11 of 19 2003]. If this is true the rock 87Sr/86Sr should vary systematically with depth in the crust because as a hydrothermal fluid passes through the crust it will progressively equilibrate with the crust and lead to progressively smaller changes in the rock compo- sition. Thus the variation in extent of alteration with depth provides further information about the process of fluid-rock reaction. This kind of one- dimensional fluid flow has been modeled for the oceanic crust under the assumption of kinetically hindered fluid-rock equilibration [Bickle and Teagle, 1992; Teagle et al., 2003]. We follow this approach here using the model of Bickle [1992; cf. Blattner and Lassey, 1989] in which the rate of exchange of Sr-isotopes between the fluid and rock is assumed to be linearly proportional to the difference in 87Sr/86Sr between the rock and fluid at all depths in the system (i.e., linear kinetics). The concentration of Sr in the fluid and rock are assumed fixed in this model. In this linear kinetic model the relative rates of advection of Sr through the system by the fluid versus Sr-isotope exchange with the wall rock is controlled by a dimensionless Damköhler number (ND), with the ND increasing with increasing exchange rate. [33] Since this linear kinetic model includes mass balance, if the fluid at the base of the system is fixed to equal the vent fluid composition, assuming no fluid-rock reaction during discharge, identical time-integrated fluid fluxes are computed using this model as using the fluid/rock ratio. As an alternative approach to modeling the fluid flux we ignore the vent fluid composition and fit the rock data using this model. Using the same input parameters as previous studies (rock density is 2.7 kg m3; fluid density is 1 kg m3 [e.g., Teagle et al., 2003]) the lack of significant variation in rock 87Sr/86Sr with depth requires a slow rate of Sr- isotope exchange that is approximated by a ND = 0.1. We kept this constant in all model runs for two reasons. First the lack of significant variation in the rock composition with depth in the crust means that the rock 87Sr/86Sr provides a poor constraint on this parameter. Second, keeping this value constant facilitates comparison of the time-integrat- ed fluid fluxes between locations. The best fit of the model to the data was assessed by minimization Table 2. Estimated Fluid Fluxes for Pito Deep, Hess Deep, and ODP Hole 504B Modela Pito Deep Hess Deep 504B Pito Deep Area A Area B Fluid/rock mass balance 2.6  106 1.5  106 2.0  106 2.4  106 2.6  106 Uncertainty ±0.9  106 ±0.9  106 ±0.6  106 ±0.3  106 ±0.6  106 Linear kinetics approximationb 2.4  106 (0.1) 1.04  106 (0.1) 2.1  106 (0.1) 2.2  106 (0.1) 2.3  106 (0.1) Standard error of best fit 4.6  108 9.5  105 3.9  104 4.7  108 3.4  108 Predicted 87Sr/86Sr of basal fluid 0.7040 0.7034 0.7040 0.7044 0.7039 Published fluid flux 3.2  106 (0.07) 1.7  106 (0.2); 5  106 (1) Reference Gillis et al. [2005] Teagle et al. [2003]; Bickle and Teagle [1992] Diffusive exchange around a ‘‘single crack’’b 5.4  106 (0.13) 3.6  106 (0.09) 6.4  106 (0.15) 4.6  106 (0.12) 5.8  106 (0.14) Standard error of best fit 4.4  109 8.2  109 3.1  104 2.7  108 3.3  109 Predicted 87Sr/86Sr of basal fluid 0.7072 0.7074 0.7076 0.7070 0.7072 a Fluid fluxes are in kg m2. Model input parameters: densities of 2.7 kg m3 and 1 kg m3 in rock and seawater, respectively, porosity of 0.02 kg m3. Location-specific parameters for ODP Hole 504B; seawater Sr concentration of 8 ppm, with modified 87Sr/86Sr of 0.7084 due to reaction with lava pile for ODP Hole 504B [Teagle et al., 2003], initial Sr concentrations of sheeted dike complex are 60 ppm, 87Sr/86Sr of fresh rock is 0.70245 [Verma and Schilling, 1982; Schilling et al., 2003]. Pito Deep has seawater Sr concentration of 8 ppm and 87Sr/86Sr 0.7086, initial Sr concentrations of sheeted dike complex are 95 ppm (100 ppm for Area A and 95 ppm for Area B), and 87Sr/86Sr of fresh rock is 0.70239. Hess Deep has seawater Sr concentration of 8 ppm and 87Sr/86Sr 0.7086, initial Sr concentrations of sheeted dike complex are 90 ppm, and 87Sr/86Sr of fresh rock is 0.70245 [Gillis et al., 2005; this study]. b Damköhler number used is given in parentheses. Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 12 of 19 of the standard error of difference between the model and the median rock 87Sr/86Sr at 100 m depth intervals. The median was chosen to lessen the bias toward outliers observed at high 87Sr/86Sr, especially at ODP Hole 504B. [34] Using this approach, the time-integrated fluid fluxes calculated for Pito Deep and ODP Hole 504B are similar to those calculated using the closed-system fluid/rock ratio (Figure 7 and Table 2). The calculated 87Sr/86Sr for the fluid at the base of the system is comparable to vent fluids from the East Pacific and quartz-epidote veins from Pito Deep and ODP Hole 504B. The fluid flux calculated using this approach for the sheeted dike complex at Hess Deep is slightly lower than that calculated using a closed system fluid/rock ratio because the resulting fluid has a lower 87Sr/86Sr than the average vent fluid (Figure 7). As discussed above, leaching of Sr from the sheeted dikes can lower the 87Sr/86Sr of resulting fluid compositions meaning that the computed fluid 87Sr/86Sr in all of these models are maximums. [35] Our estimate for time-integrated fluid fluxes in the sheeted dike complex at ODP Hole 504B is similar to the fluid flux calculated using a ND of 0.2 and fitting the model fluid 87Sr/86Sr to the measured 87Sr/86Sr of anhydrite assuming these anhydrite formed in equilibrium with recharging fluids [Teagle et al., 2003]. The difference between the estimated fluid flux for Hess Deep and the published value [Gillis et al., 2005] is due to the higher exchange rate (ND) used here. [36] The misfit between the models and data pro- vides insight into whether Sr-isotope exchange occurs solely during recharge as assumed in this Figure 7. Linear kinetic model for fluid fluxes through sheeted dike complexes at (a) Pito Deep data with median ± 95% confidence interval for 100 m depth intervals, (b) Hess Deep with median ± 95% confidence interval for 100 m depth intervals, and (c) ODP Hole 504B with median ± 95% confidence interval for 100 m depth intervals. Model input parameters: ND = 0.1 in all locations, densities of 2.7 kg m 3 and 1 kg m3 in rock and seawater, respectively [Teagle et al., 2003], porosity of 0.02 kg m3. Location-specific parameters for ODP Hole 504B; seawater Sr concentration of 8 ppm, with modified 87Sr/86Sr of 0.7084 due to reaction with lava pile for ODP Hole 504B [Teagle et al., 2003]. Initial Sr concentrations of sheeted dike complex are 60 ppm; 87Sr/86Sr of fresh rock is 0.70245 [Verma and Schilling, 1982; Schilling et al., 2003]. Pito Deep has seawater Sr concentration of 8 ppm and 87Sr/86Sr 0.7086, initial Sr concentrations of sheeted dike complex are 95 ppm (100 ppm for Area A and 95 ppm for Area B), and 87Sr/86Sr of fresh rock is 0.70239. Hess Deep has seawater Sr concentration of 8 ppm and 87Sr/86Sr 0.7086, initial Sr concentrations of sheeted dike complex are 90 ppm, and 87Sr/86Sr of fresh rock is 0.70245 [Gillis et al., 2005; this study]. Final fluid at base of sheeted dike complex is assumed to discharge without further modification to form vent fluids. Calculated final fluid compositions are within the range of 87Sr/86Sr for vent fluids and the 87Sr/86Sr composition of quartz-epidote veins at Pito Deep and ODP Hole 504B. Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 13 of 19 modeling. If the misfits were randomly distributed with depth then this model would be consistent with the data but this is not the case (Figure 8a). At Pito Deep the rocks are, on average, less altered than the model predicts in the top 250 m of the sheeted dikes and more altered than the model predicts at the very base of the sheeted dike complex. Unfortunately we have few samples from the base of the sheeted dike complex adding to the uncertainty in this conclusion. For Hess Deep the smaller depth range of sampling, and the paucity of samples from deep in the sheeted dike complex, makes it even more difficult to determine how well the model fits the data. The model fits the data reasonably well in the upper portion of the sheeted dike complex but samples from the base of the sheeted dike complex have higher 87Sr/86Sr than predicted by the model. The much larger data set, especially in the deeper dikes, for ODP Hole 504B allows much more rigorous testing of the model. For this sample suite the upper dikes are less altered than the model predicts and the deeper dikes are more altered than the model predicts (Figure 8a). Using a ND of 0.2 to model the data from ODP Hole 504B as proposed by Teagle et al. [2003] accentuates these systematic misfits. These systematic misfits suggest that the basic premise of this model, that fluid-rock reaction occurs homo- geneously during one-dimensional fluid transport, is not valid. [37] One assumption of the modeling presented in Figure 7 is that the rate of Sr-isotopic exchange between the fluid and rock is assumed simply to be linearly proportional to the difference in fluid and rock 87Sr/86Sr ratio. A more realistic model might be diffusive exchange of Sr from fluid filled cracks into the wall rock. Bickle [1992] shows that for high ND, the linear kinetic and diffusive models give the same result but at low ND (<1) the effect on the rock composition is significantly different. To test whether this can explain the misfit between the model and data we apply this model in the following section. 5.3.3. Low Rates of Fluid-Rock Reaction: Diffusive Exchange Around a ‘‘Single Crack’’ [38] If the fluid-rock reaction process leads to the rate of Sr exchange not being linearly dependent on the difference in 87Sr/86Sr between the rock and fluid, but instead being slower, this could poten- tially explain the lack of any systematic change in Figure 8. Deviation between samples and model curves for sheeted dikes at Pito Deep, Hess Deep, and ODP Hole 504B. (a) Linear kinetic model ND = 0.1. Inset shows deviations between median and the linear kinetic model for 100 m depth intervals. (b) Single crack model. Inset shows deviations between median and the single crack model for 100 m depth intervals. Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 14 of 19 rock composition with depth. An example of this kind of exchange mechanism is if fluid-rock reac- tion occurs around fractures where the alteration haloes around these fractures do not overlap. Tang et al. [1981] present an analytical solution to the equations describing the concentration of a tracer in the wall rock around a fracture through which fluid flows (i.e., fluid flow perpendicular to Sr- exchange). Bickle [1992] integrated this to obtain a solution (his equation (38)) for the average rock composition as a function of a ND equivalent to that discussed above. Bickle [1992] suggests that this model is more appropriate for tracing fluid flow when the fluid-rock exchange rates are low (ND < 0.3). We have modeled the variation in 87Sr/86Sr in the sheeted dikes with depth below the lava-sheeted dike transition in this way for all three areas. We allow the ND to vary as well as the fluid flux and calculated best fitting fluid fluxes and ND (in parentheses) of 5.4  106 kg m2 (0.13), 3.6  106 kg m2 (0.09) and 6.4  106 kg m2 (0.15) for Pito Deep, Hess Deep and ODP Hole 504B, respectively (Figure 9 and Table 2). [39] These models fit the data better than the models presented above with little systematic change in the misfit between the model and data with depth in the sheeted dike complex (Figure 8b). However, the fluid fluxes calculated using this approach are significantly greater than those cal- culated above leading to the fluid becoming less rock-dominated. The fluid 87Sr/86Sr at the base of the system can be calculated by mass balance to be between 0.7070 and 0.7076, substantially higher than vent fluids. Assuming these final fluid com- positions are vented without further modification during discharge they are far too high in 87Sr/86Sr for this model to provide an accurate representation of the behavior of Sr in the axial hydrothermal system. Even leaching a small amount of Sr from the crust (Figures 4 and 5) cannot bring the fluid 87Sr/86Sr down to the range of vent fluids. Thus, unless there is a region of the crust from which Sr is leached substantially more than from any ob- served samples this model appears unsatisfactory. 5.4. Fluid-Rock Reaction During Discharge as Well as Recharge? [40] Since neither pervasive, kinetically inhibited, fluid-rock reaction nor fluid-rock reaction around isolated channels appears to fit both the variation in rock composition with depth and the vent fluids compositions, we consider a third model in which fluid-rock reaction occurs during both recharge and discharge. This might be expected since the fluid is Figure 9. Single crack model of fluid flux at (a) Pito Deep with data for the sheeted dikes and median ± 95% confidence interval for 100 m depth intervals, (b) Hess Deep with data for the sheeted dikes and median ± 95% confidence interval for 100 m depth intervals, and (c) ODP Hole 504B with data for the sheeted dikes and median ± 95% confidence interval for 100 m depth intervals. See Table 2 for ND and fluid fluxes. Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 15 of 19 likely to be more reactive once it is heated near the base of the system [e.g., Fontaine and Wilcock, 2007; Coogan, 2008]. Because this model has more free parameters our aim is not to fit the data perfectly but rather to test whether the model is broadly consistent with the data. We use a perva- sive flow model with kinetically limited exchange (linear kinetics) and assume that recharge and discharge of the fluid occur over the same part of the sheeted dike complex. Realistically, recharge and discharge would be expected to occur at different times in the axial history of the ocean crust, with recirculating hydrothermal fluids poten- tially affecting any profile of the ocean crust multiple times [Fontaine and Wilcock, 2007]. The model is constructed in two parts with the resulting fluid and rock Sr isotopic compositions from the recharging leg providing input parameters for the discharging leg and alteration of the sheeted dike complex occurring during discharge being additive to that accrued during recharge. On the basis of the idea that the upflowing fluid will be hotter and hence exchange isotopes with the crust more rap- idly we use a ND = 0.05 in the recharge zone and ND = 0.08 in the upflow zone. Fitting the data from Pito Deep (Figure 10) using a similar fluid flux as in the unidirectional linear-kinetic model gives a vent fluid of 87Sr/86Sr 0.70352 and matches the 87Sr/86Sr of vein hosted epidote at 220 and 370 m, that are assumed to have formed from discharging fluids. This modeling demonstrates that if fluid- rock reaction occurs within the regions of upflow in addition to those of downflow there is no requirement for there to be any variation of 87Sr/86Sr with depth in the sheeted dike complex unlike if reaction occurs only in the recharge zone. [41] Thus far, limited 87Sr/86Sr data for fluid de- rived hydrothermal minerals (0.7034–0.7041 [Teagle et al., 1998b; this study]) indicate precip- itation from highly evolved fluids that would be expected to be associated with hot reactive fluids deep in the hydrothermal system or upflow zones analogous to the discharging leg of this model. There is no clear evidence for reaction during recharge of cooler fluids, perhaps due to limited fluid-rock reaction or overprinting. 6. Discussion and Conclusions [42] There are three important conclusions from this study: [43] 1. The remarkable similarity in the Sr-isotope profile recorded by the sheeted dike complex in three different areas suggests that fluid flow through the sheeted dike complex is surprisingly uniform. This is despite significant differences in the spreading rate and thickness of the sheeted dike complex. Additionally there are differences in the protolith composition, in particular plagioclase An-content, which might be expected to lead to difference in the reactivity of the crust but do not appear to be important (Table 1). Neither does the extent of mineralogical alteration or dominant secondary phase appear critical (Table 1). [44] 2. The modification of the 87Sr/86Sr in both the sheeted dike complex and vent fluids are consistent with minimum fluid fluxes through the sheeted dike complex at intermediate to fast spreading ridges of between 1.5 and 2.6  106 kg m2. Evidence for leaching of Sr from the crust into the hydrothermal fluid means that this is an underes- timate of the true flux by between 0.2 and 0.4 kg m2. These fluxes ignore two potentially important factors. First, if significant anhydrite is precipitated in the recharge zone this could deplete the recharg- ing fluid of Sr leading to a higher fluid flux being required for a given change in rock composition. Second, if a substantial portion of the fluid flux Figure 10. Downwelling and upwelling of fluid modeled by linear kinetic exchange for sheeted dikes at Pito Deep with median ± 95% confidence interval for 100 m depth intervals. Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 16 of 19 through the axial system is channelized this will not be recorded by the samples studied. [45] 3. The lack of a significant change in the 87Sr/86Sr of samples from the sheeted dike com- plex with depth in the crust is inconsistent with all Sr-isotope exchange occurring within a recharge zone. This can be explained if fluid-rock reaction also occurs within the discharge zone. Acknowledgments [46] Journal reviews from Damon Teagle and Mike Bickle are gratefully acknowledged. We would like to thank Jeff Karson and Emily Klein for the opportunity to work on the Pito Deep project. We are grateful to Jody Spence at the University of Victoria and Bruno Kieffer at the Pacific Centre for Isotopic and Geochemical Research, University of British Columbia, for analytical assistance and Mike Bickle and Nick Hayman for discussion. We thank the Pito Deep shipboard scientific party for collecting and logging samples and ac- knowledge NSF grant OCE-0222154 to J. A. Karson and E. M. Klein that funded the cruise. Postcruise science was funded through the NSERC SRO program. References Albarède, F., A. Michard, J. F. Minster, and G. Michard (1981), 87Sr/86Sr ratios in hydrothermal waters and deposits from the East Pacific Rise at 21N, Earth Planet. Sci. Lett., 55(2), 229–236, doi:10.1016/0012-821X(81)90102-3. Alt, J. C. (1995). Subseafloor processes in mid-ocean ridge hydrothermal systems, in Seafloor Hydrothermal Systems: Physical, Chemical, Biological and Geological Interactions, Geophys. Monogr. Ser., vol. 91, edited by S. E. Humphris et al., pp. 85–114, AGU, Washington, D.C. Alt, J. C., T. F. Anderson, L. Bonnell, and K. Muehlenbachs (1989), Mineralogy, chemistry and stable isotope composi- tions of hydrothermally altered sheeted dikes: ODP Hole 504B, Leg 111, Proc. Ocean Drill. Program Sci. Results, 111, 27–40. Alt, J. C., D. A. H. Teagle, W. Bach, A. N. Halliday, and J. Erzinger (1996a), Stable and strontium isotopic profiles through hydrothermally altered upper oceanic crust, Hole 504B, Proc. Ocean Drill. Program Sci. Results, 148, 57–69. Alt, J. C., C. Laverne, D. Vanko, P. Tartarotti, D. A. H. Teagle, W. Bach, E. Zulegar, J. Erzinger, and J. Honnorez (1996b), Hydrothermal alteration of a section of upper oceanic crust in the eastern equatorial Pacific: A synthesis of results from DSDP/ODP Legs 69, 70, 83, 111, 137, 140, 148 at Site 504B, Proc. Ocean Drill. Program Sci. Results, 148, 417– 434. Bach, W., and S. E. Humphris (1999), Relationship between the Sr and O isotope compositions of hydrothermal fluids and the spreading and magma-supply rates at oceanic spread- ing centers, Geology, 27, 1067–1070, doi:10.1130/0091- 7613(1999)027<1067:RBTSAO>2.3.CO;2. Bach, W., B. Peucker-Ehrenbrink, S. R. Hart, and J. S. Blusztajn (2003), Geochemistry of hydrothermally altered oceanic crust: DSDP/ODP Hole 504B - Implications for seawater-crust ex- change budgets and Sr- and Pb-isotopic evolution of the man- tle, Geochem. Geophys. Geosyst., 4(3), 8904, doi:10.1029/ 2002GC000419. Bau, M. (1991), Rare-earth element mobility during hydrother- mal and metamorphic fluid-rock interaction and the signifi- cance of the oxidation state of europium, Chem. Geol., 93, 219–230, doi:10.1016/0009-2541(91)90115-8. Berndt, M. E., W. E. Seyfried, and J. W. Beck (1988), Hydro- thermal alteration processes at midocean ridges: Experimen- tal and theoretical constraints from Ca and Sr exchange reactions and Sr isotopic ratios, J. Geophys. Res., 93(B5), 4573–4583, doi:10.1029/JB093iB05p04573. Bickle, M. J. (1992), Transport mechanisms by fluid-flow in metamorphic rocks: Oxygen and strontium decoupling in the Trois Seigneurs massif-A consequence of kinetic disper- sion?, Am. J. Sci., 292, 289–316. Bickle, M. J., and D. A. H. Teagle (1992), Strontium alteration in the Troodos ophiolite: Implications for fluid fluxes and geochemical transport in mid-ocean ridge hydrothermal sys- temsEarth Planet. Sci. Lett., 113, 219–237, doi:10.1016/ 0012-821X(92)90221-G. Bickle, M. J., D. A. H. Teagle, J. Beynon, and H. J. Chapman (1998), The structure and controls on fluid-rock interactions in ocean ridge hydrothermal systems: Constraints from the Troodos ophiolite, in Modern Ocean Floor Processes and the Geological Record, edited by R. A. Mills and K. Harrison, Geol. Soc. Spec. Publ., 148, 127–152. Blattner, P., and K. Lassey (1989), Stable-isotope exchange fronts, Damköhler numbers, and fluid to rock ratios, Chem. Geol., 78, 381–392, doi:10.1016/0009-2541(89)90070-3. Butterfield, D. A., and G. J. Massoth (1994), Geochemistry of north Cleft segment vent fluids: Temporal changes in chlori- nity and their possible relation to recent volcanism, J. Geo- phys. Res., 99(B3), 4951–4968, doi:10.1029/93JB02798. Butterfield, D. A., R. E. McDuff, M. J. Mottl, M. D. Lilley, J. E. Lupton, and G. J. Massoth (1994), Gradients in the composition of hydrothermal fluids from the Endeavour seg- ment vent field: Phase separation and brine loss, J. Geophys. Res., 99(B5), 9561–9583, doi:10.1029/93JB03132. Charlou, J. L., Y. Fouquet, J. P. Danval, and J. M. Auzende (1996), Mineral and gas chemistry of hydrothermal fluids on an ultrafast spreading ridge: East Pacific Rise, 17 to 19S (Naudar cruise, 1993) phase separation processes controlled by volcanic and tectonic activity, J. Geophys. Res., 101(B7), 15,899–15,919, doi:10.1029/96JB00880. Coogan, L. A. (2008), Reconciling temperatures of meta- morphism, fluid fluxes, and heat transport in the upper crust at intermediate to fast spreading mid-ocean ridges, Geochem. Geophys. Geosyst., 9, Q02013, doi:10.1029/2007GC001787. Coogan, L. A., K. M. Gillis, C. J. MacLeod, G. M. Thompson, and R. Hékinian (2002), Petrology and geochemistry of the lower ocean crust formed at the East Pacific Rise and exposed at Hess Deep: A synthesis and new results, Geochem. Geophys. Geosyst., 3(11), 8604, doi:10.1029/ 2001GC000230. Davis, A. C., M. J. Bickle, and D. A. H. Teagle (2003), Im- balance in the oceanic strontium budget, Earth Planet. Sci. Lett., 211, 173–187, doi:10.1016/S0012-821X(03)00191-2. Eggins, S. M., J. D. Woodhead, L. P. J. Kinsley, G. E. Morti- mer, P. Sylvester, M. T. McCulloch, J. M. Hergt, and M. R. Handler (1997), A simple method for the precise determina- tion of 40 trace elements in geological samples by ICPMS using enriched isotope internal standardization, Chem. Geol., 134, 311–326, doi:10.1016/S0009-2541(96)00100-3. Fontaine, F. J., and W. S. D. Wilcock (2007), Two-dimensional numerical models of open-top hydrothermal convection at high Rayleigh and Nusselt numbers: Implications for mid- Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 17 of 19 ocean ridge hydrothermal circulation, Geochem. Geophys. Geosyst., 8, Q07010, doi:10.1029/2007GC001601. Francheteau, J., P. Patriat, J. Segoufin, R. Armijo, M. Doucoure, A. Yelles-Chaouche, J. Zukin, S. Calmant, D. F. Naar, and R. C. Searle (1988), Pito and Orongo fracture zones: The northern and southern boundaries of the Easter microplate (southeast Pacific), Earth Planet. Sci. Lett., 89, 363–374, doi:10.1016/0012-821X(88)90123-9. Francheteau, J., R. Armijo, J. L. Cheminée, R. Hékinian, P. Lonsdale, and N. Blum (1990), 1 Ma East Pacific Rise oceanic crust and uppermost mantle exposed by rifting in Hess Deep (equatorial Pacific Ocean), Earth Planet. Sci. Lett., 101, 281–295, doi:10.1016/0012-821X(90)90160-Y. Gillis, K. M. (1995), Controls on hydrothermal alteration in a section of fast-spreading oceanic crust, Earth Planet. Sci. Lett., 134, 473–489, doi:10.1016/0012-821X(95)00137-2. Gillis, K. M., K. Muehlenbachs, M. Stewart, T. Gleeson, and J. Karson (2001), Fluid flow patterns in fast spreading East Pacific Rise crust exposed at Hess Deep, J. Geophys. Res., 106(B11), 26,311–26,329, doi:10.1029/2000JB000038. Gillis, K. M., L. A. Coogan, and R. Pedersen (2005), Stron- tium isotope constraints in fluid flow in the upper oceanic crust at the East Pacific Rise, Earth Planet. Sci. Lett., 232, 83–94, doi:10.1016/j.epsl.2005.01.008. Heft, K. L., K. M. Gillis, M. A. Pollock, J. A. Karson, and E. M. Klein (2008), Role of upwelling hydrothermal fluids in the development of alteration patterns at fast spreading ridges: Evidence from the sheeted dike complex at Pito Deep, Geochem. Geophys. Geosyst., 9, Q05O07, doi:10.1029/2007GC001926. Hékinian, R., J. Francheteau, R. Armijo, J. P. Cogne, M. Constanti, J. Girardeau, R. Hey, D. F. Naar, and R. Searle (1996), Petrology of the Easter microplate region in the South Pacific, J. Volcanol. Geotherm. Res., 72, 259–289, doi:10.1016/0377-0273(96)00013-3. Hess, J., M. L. Bender, and J.-G. Schilling (1986), Evolution of the ratio of strontium-87 to strontium-86 in seawater from Cretaceous to present, Science, 231, 979–984, doi:10.1126/ science.231.4741.979. Hey, R., G. L. Johnson, and A. Lowrie (1977), Recent plate mo- tions in the Galapagos area, Geol. Soc. Am. Bull., 88, 1385– 1403, doi:10.1130/00167606(1977)88<1385:RPMITG> 2.0.CO;2. Hey, R. N., P. D. Johnson, F. Martinez, J. Korenga, M. L. Somers, Q. J. Huggett, T. P. LeBas, R. I. Rusby, and D. F. Naar (1995), Plate boundary reorganization at a large-offset, rapidly propagating rift, Nature, 378, 167–170, doi:10.1038/ 378167a0. Hey, R. N., et al. (2002), Preliminary attempt to characterize the rotation of seafloor in the Pito Deep area of the Easter Microplate using a submersible magnetometer, Mar. Geo- phys. Res., 23, 1–12, doi:10.1023/A:1021257915420. Hilgen, F. J., W. Krijgsman, C. G. Langereis, L. J. Lourens, A. Santarelli, and W. J. Zachariasse (1995), Extending the astronomical (polarity) time scale into the Miocene, Earth Planet. Sci. Lett., 136, 495–510, doi:10.1016/0012- 821X(95)00207-S. Hinkley, T. K., and M. Tatsumoto (1987), Metals and isotopes in Juan de Fuca ridge hydrothermal fluids and their asso- ciated solid materials, J. Geophys. Res., 92(B11), 11,400– 11,410, doi:10.1029/JB092iB11p11400. Hofmann, A. W., and W. M. White (1982), Mantle plumes from ancient oceanic crust, Earth Planet. Sci. Lett., 57, 421–436, doi:10.1016/0012-821X(82)90161-3. Karson, J. A. and Pito Deep Shipboard Scientific Party (2005), Pito Deep cruise report, Dep. of Earth Sci., Syracuse Univ., Syracuse, N.Y. Karson, J. A., et al. (2002), Structure of uppermost fast-spread oceanic crust exposed at the Hess Deep Rift: Implications for subaxial processes at the East Pacific Rise, Geochem. Geo- phys. Geosyst., 3(1), 1002, doi:10.1029/2001GC000155. Kusakabe, M., T. Shibata, M. Yamamoto, S. Mayeda, H. Kagami, H. Honma, H. Masuda, and H. Sakai (1989), Petrology and isotope characteristics (H, O, S, Sr, and Nd) of basalts from Ocean Drilling program Hole 504B, Proc. Ocean Drill. Program Sci. Results, 111, 47–60. Laverne, C., D. A. Vanko, P. Tartarotti, and J. C. Alt (1995), Chemistry and geothermometry of secondary minerals from the deep sheeted dike complex, DSDP/ODP Hole 504B, Proc. Ocean Drill. Program Sci. Results, 137/140, 167–189. Laverne, C., P. Agrinier, D. Hermitte, and M. Bohn (2001), Chemical fluxes during hydrothermal alteration of a 1200-m long section of dikes in the oceanic crust, DSDP/ODP Hole 504B, Chem. Geol., 181, 73–98, doi:10.1016/S0009- 2541(01)00277-7. Lonsdale, P. (1988), Structural pattern of the Galapagos Microplate and evolution of the Galapagos Triple Junction, J. Geophys. Res., 93(B11), 13,551–13,574, doi:10.1029/ JB093iB11p13551. Mahoney, J. J., J. M. Sinton, M. D. Kurz, J. D. Macdougall, K. J. Spencer, and G. W. Lugmair (1994), Isotope and trace element characteristics of a super-fast spreading ridge: East Pacific rise, 13–23S, Earth Planet. Sci. Lett., 121, 173– 193, doi:10.1016/0012-821X(94)90039-6. Martinez, F., D. F. Naar, T. B. Reed, and R. N. Hey (1991), Three-dimensional SEAMARC II, gravity and magnetics study of large-offset rift propagation at the Pito Rift, Easter Microplate, Mar. Geophys. Res., 13, 255–285, doi:10.1007/ BF00366279. McKay, G., L. Le, J. Wagstaff, and G. Crozaz (1994), Experi- mental partitioning of rare earth elements and strontium: Constraints on petrogenesis and redox conditions during crystallization of Antarctic angrite Lewis Cliff 86010, Geo- chim. Cosmochim. Acta, 58, 2911–2919, doi:10.1016/0016- 7037(94)90124-4. Merlivat, L., F. Pineau, and M. Javoy (1987), Hydrothermal vent waters at 13N on the East Pacific Rise: Isotopic com- position and gas concentration, Earth Planet. Sci. Lett., 84, 100–108, doi:10.1016/0012-821X(87)90180-4. Palmer, M. R. (1992), Controls over the chloride concentration of submarine hydrothermal vent fluids: Evidence from Sr/Ca and 87Sr/86Sr ratios, Earth Planet. Sci. Lett., 109, 37–46, doi:10.1016/0012-821X(92)90072-4. Palmer, M. R., and J. M. Edmond (1989), The strontium iso- tope budget of the modern ocean, Earth Planet. Sci. Lett., 92, 11–26, doi:10.1016/0012-821X(89)90017-4. Perk, N. W., L. A. Coogan, J. A. Karson, E. M. Klein, and H. D. Hanna (2007), Petrology and geochemistry of primi- tive lower oceanic crust from Pito Deep: Implications for the accretion of the lower crust at the southern East Pacific Rise, Contrib. Mineral. Petrol., 154, doi:10.1007/s00410-007- 0210-z. Piepgras, D. J., and G. J. Wasserburg (1985), Strontium and neodymium isotopes in hot springs on the East Pacific Rise and Guaymas Basin, Earth Planet. Sci. Lett., 72, 341–356, doi:10.1016/0012-821X(85)90057-3. Schiffman, P., B. M. Smith, R. J. Varga, and E. M. Moores (1987), Geometry, conditions and timing of off-axis hydro- Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 18 of 19 thermal metamorphism and ore-deposition in the Solea gra- ben, Nature, 325, 423–425, doi:10.1038/325423a0. Schilling, J. G., D. Fontignie, J. Blichert-Toft, R. H. Kingsley, and U. Tomza (2003), Pb-Hf-Nd-Sr isotope variations along the Galapagos spreading center (101-83W): Constraints on the dispersal of the Galapagos mantle plume, Geochem. Geo- phys. Geosyst., 4(10), 8512, doi:10.1029/2002GC000495. Seyfried, W. E. (1987), Experimental and theoretical con- straints on hydrothermal alteration processes at mid-ocean ridges, Annu. Rev. Earth Sci., 15, 317–335, doi:10.1146/ annurev.ea.15.050187.001533. Tang, D. H., E. O. Frind, and E. A. Sudickey (1981), Contam- ination transport in fractured porous media: analytical solu- tion for a single fracture, Water Resour. Res., 17, 555–564, doi:10.1029/WR017i003p00555. Taylor, H. P. (1977), Water/rock interaction and the origin of H2O in granitic batholiths, J. Geol. Soc., 133(6), 509–558, doi:10.1144/gsjgs.133.6.0509. Teagle, D. A. H., J. C. Alt, and A. N. Halliday (1998a), Tra- cing the chemical evolution of fluids during hydrothermal recharge: Constraints from anhydrite recovered in ODP Hole 504B, Earth Planet. Sci. Lett., 155, 167–182, doi:10.1016/ S0012-821X(97)00209-4. Teagle, D. A. H., J. C. Alt, and A. N. Halliday (1998b), Tra- cing the chemical evolution of fluids in the upper oceanic crust: Sr isotopic constraints from DSDP/ODP Holes 504B and 896A, in Modern Ocean Floor Processes and the Geo- logical Record, edited by R. A. Mills and K. Harrison, Geol. Soc. Spec. Publ., 148, 81–97. Teagle, D. A. H., M. J. Bickle, and J. C. Alt (2003), Recharge flux to ocean-ridge black smoker systems: a geochemical estimate from ODP Hole 504B, Earth Planet. Sci. Lett., 210, 81–89, doi:10.1016/S0012-821X(03)00126-2. Vanko, D. A., C. Laverne, P. Tartorotti, and J. C. Alt (1996), Chemistry and origin of secondary minerals from, the deep sheeted dikes cored during ODP Leg 148, Hole 504B, Proc. Ocean Drill. Program Sci. Results, 148, 71–86. Verma, S. P., and J. G. Schilling (1982), Galapagos Hotspot spreading center system: 2. 87Sr/86Sr and large ion lithophile element variations (85W–101E), J. Geophys. Res., 87, 10,838–10,856, doi:10.1029/JB087iB13p10838. Von Damm, K. L. (1990), Seafloor hydrothermal activity: Black smoker chemistry and chimneys, Annu. Rev. Earth Planet. Sci. , 18 , 173 – 204, doi:10.1146/annure- v.ea.18.050190.001133. Von Damm, K. L. (1995), Controls on the chemistry and tem- poral variability of seafloor hydrothermal fluids, in Seafloor Hydrothermal Systems: Physical, Chemical, Biological and Geological Interactions, Geophys. Monogr. Ser., vol. 91, edited by S. E. Humphris et al., pp. 222–247, AGU, Washington, D.C. Von Damm, K. L. (2000), Chemistry of hydrothermal vent fluids from 9–10N, East Pacific Rise: ‘‘Time zero,’’ the immediate posteruptive period, J. Geophys. Res., 105(B5), 11,203–11,222, doi:10.1029/1999JB900414. Weis, D., et al. (2006), High-precision isotopic characterization of USGS reference materials by TIMS and MC-ICP-MS, Geochem. Geophys. Geosyst., 7, Q08006, doi:10.1029/ 2006GC001283. Geochemistry Geophysics Geosystems G3 barker et al.: strontium isotope constraints, 1 10.1029/2007GC001901 19 of 19"@en ; edm:hasType "Article"@en ; edm:isShownAt "10.14288/1.0075968"@en ; dcterms:language "eng"@en ; ns0:peerReviewStatus "Reviewed"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "American Geophysical Union"@en ; ns0:publisherDOI "10.1029/2007GC001901"@en ; dcterms:rights "Attribution-NonCommercial-NoDerivatives 4.0 International"@en ; ns0:rightsURI "http://creativecommons.org/licenses/by-nc-nd/4.0/"@en ; ns0:scholarLevel "Faculty"@en ; dcterms:subject "High-temperature Geochemistry"@en ; dcterms:title "Strontium isotope constraints on fluid flow in the sheeted dike complex of fast spreading crust: pervasive fluid flow at Pito Deep."@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/39090"@en .