BACKGROUND

Geological Setting

The Endeavour segment of the Juan de Fuca Ridge (JFR) generates lithosphere west of North America at ~3 cm/y (Davis and Currie, 1993; Johnson and Holmes, 1989). Topographic relief produces barriers to turbidites from the continental margin, resulting in the accumulation of sediment and burial of the eastern flank of the JFR within a region known as Cascadia Basin (Figs. F1, F2). Sedimentation rates were very high in Cascadia Basin during Pleistocene sea level low-stands, when the continental shelf was largely exposed and rivers and estuaries delivered large sediment fluxes directly to the deep ocean (Davis, Fisher, Firth, et al., 1997; Underwood et al., in press). This resulted in burial of oceanic basement rocks under thick sediments throughout much of the basin at an unusually young age. Oceanic basement is exposed to the west, where the crust is very young, and the sedimented seafloor is relatively flat to the east, except over (relatively rare) seamounts and other outcrops found near the eastern end of the Leg 168 transect (Figs. F1, F3). Basement relief is dominated by linear ridges and troughs oriented subparallel to the spreading center and produced mainly by faulting, variations in magmatic supply at the ridge, and off-axis volcanism (Davis and Currie, 1993; Kappel and Ryan, 1986). Basement relief is relatively low (–200 m) near the ridge and higher (–700 m) to the east. Low-permeability sediment limits advective heat loss across most of the ridge flank, leading to strong thermal, chemical, and alteration gradients in basement.

The study area contains structural features common to most ridge flanks: extrusive igneous basement overlain by sediments, abyssal hill topography, high-angle faulting, and basement outcrops. Although the work sites may not be typical of all ridge-flank settings (higher than average sedimentation rate, younger buried basement, and stronger lateral gradients in temperature), the field area is ideal, in part because of these extreme conditions. High gradients result in strong signals that rise above natural and experimental noise. The high sedimentation rate allows us to work on crust that is much younger than we could study otherwise, providing indications of ridge-crest as well as ridge-flank properties and allowing study of sites in different hydrologic settings that are close together. Because many experiments have been completed in this area (seismic, thermal, geochemical, and surface/borehole), we can "calibrate" and compare interpretations based on different methods.

Marine geophysical surveys in this region began in the 1950s and 1960s, but the first detailed studies of this ridge flank intended to resolve the existence and influence of hydrothermal circulation were completed in mid- to late 1980s (Davis et al., 1989, 1992; Mottl and Wheat, 1994; Rohr, 1994). These studies included single- and multichannel seismic, gravity, magnetics, heat flow, and coring (with associated sediment and pore fluid analyses). Becker et al. (2000) show results from a 1992 John P. Tully survey that collected seismic data in the Second Ridge area, with an emphasis on nearby basement outcrops. Seismic results from two later surveys are summarized by Rosenberger et al. (2000). Davis et al. (1997a) compiled heat flow data collected between 1978 and 1995.

Extensive site surveys in support of IODP Expedition 301 were completed by the Sonne and the Thomas G. Thompson in 2000 (ImageFlux and RetroFlux Expeditions, respectively), and an additional seismic line across the Deep Ridge sites was collected during another expedition in 2002. Example seismic lines illustrate key features of sediments and uppermost basement at IODP Expedition 301 sites (Fig. F4). The Second Ridge area is characterized by typical basement relief of 300–400 m, usually overlain by 250–600 m of sediment. To the north and south of Expedition 301 sites there are basement outcrops where basalt edifices rise above the seafloor (Davis et al., 1992; Mottl et al., 1998). Seismic data show that basement relief results in part from high-angle faults with offsets of tens to >100 m (e.g., Fig. F4A, common depth points [CDP] 570 and 600; Fig. F4B, CDPs 980 and 1020). There also appear to be at least two kinds of constructional structures on the top of basement throughout this area. First, there are small buried basement highs onto which sediment was subsequently draped. (Fig. F4A, CDP 450; Fig. F4B, CDP 840). There are also places where the uppermost basement reflector is unusually strong and flat. Becker et al. (2000) identified several such features in seismic lines crossing the Baby Bare outcrop and suggested that they were sills, and similar features appear in seismic data west of Site 1026 (Fig. F4B, CDPs 1100–1140).

The sediment section shows two main zones with distinctive seismic characteristics. The uppermost 200–300 ms two-way traveltime (TWT) include prominent subhorizontal reflectors that clearly illustrate the geometry of distributary channels for turbidites that flowed from the north. Sites 1026 and U1301 are located near the western edge of one such channel, which thickens considerably over Site 1027. There are patches of acoustically incoherent sediment within these channels, where subhorizontal layering is disrupted, and high-angle, small-offset faults that are present throughout the uppermost sediment (e.g., Fig. F4A, CDP 600). The lowermost 100–200 ms TWT of sediment is generally acoustically transparent, particularly where basement is deepest. Weak layering within this section onlaps and often pinches out against basement highs.

An 80 km transect comprising 10 sites was drilled on the eastern flank of the JFR during ODP Leg 168 (Fig. F2). These sites were organized into three main ridge-flank environments. The western end of the drilling transect spanned a transition between hydrologically open and more isolated crust, documenting lateral gradients in basement temperatures, water compositions, and crustal physical properties. The rough basement area at the eastern end of the transect included considerable basement relief, large variations in sediment thickness, and isolated outcrops. The central part of the Leg 168 transect included sites located farther from regions of known basement exposure, where sediment thickness is more uniform.

Sediments recovered during Leg 168 included mainly sandy and silty turbidites and hemipelagic mud, with carbonate-rich intervals found just above basement at most sites (Davis, Fisher, Firth, et al., 1997; Underwood et al., in press). Sediments were generally unaltered by underlying hydrothermal processes, except for relatively subtle indications close to basement (Buatier et al., 2001). Shallow basement rocks were mainly fresh to altered pillow lavas and massive flows having a tholeiitic composition, but hyaloclastite breccia and a diabase sill were recovered at Sites 1026 and 1027, respectively. The extent of alteration generally increased from west to east, along with crustal age and basement temperature, and there was no indication that basement rocks had experienced temperatures in excess of those observed during Leg 168 (Giorgetti et al., 2001; Hunter et al., 1999; Marescotti et al., 2000). Alteration minerals included clays, carbonates, zeolites, and sulfides.

Heat flow and upper basement temperatures along the Leg 168 drilling transect show several notable trends (Davis and Becker, 2002; Davis et al., 1999; Davis, Fisher, Firth, et al., 1997; Fisher et al., 1997; Pribnow et al., 2000). Heat flow values determined during Leg 168 increase over the western 20 km of the transect, from Site 1023 to Sites 1030 and 1031 (Fig. F2). These values vary from well below to well above standard reference curves for conductively cooling lithosphere (e.g., Parsons and Sclater, 1977; Stein and Stein, 1994). Heat flow in the middle and eastern end of the drilling transect is broadly consistent with reference curves, but local (sometimes large) variations in heat flow result from vigorous hydrothermal circulation within rugged basement below the seafloor (Fig. F2B). This circulation locally homogenizes uppermost basement temperatures such that seafloor heat flow patterns follow basement relief (e.g., Davis and Becker, 2002; Davis et al., 1997b; Spinelli and Fisher, 2004). Upper basement temperatures tend to increase monotonically from west to east along the drilling transect, from ~15°C at Site 1023 to ~64°C at Sites 1026 and 1027 (Fig. F2B) (Davis and Becker, 2002).

Thermal observations along the Leg 168 transect may be interpreted to indicate that the dominant direction of fluid flow is from the west to the east (Davis et al., 1999; Davis, Fisher, Firth, et al., 1997; Stein and Fisher, 2003), but pore fluid samples obtained from sediments collected immediately above basement, in combination with samples from basement boreholes and shallow sediment cores, demonstrate that this interpretation is not consistent with observations (Fig. F2C). The western end of the drilling transect shows increasing alteration from west to east, consistent with rising temperatures in upper basement, but fluid recovered from Sites 1030 and 1031 is more altered than would be predicted on the basis of inferred temperatures in upper basement. In fact, this fluid has a geochemical signature consistent with alteration at 65°–70°C, much like the fluid recovered from Sites 1026 and 1027 and from springs on Baby Bare outcrop (Elderfield et al., 1999; Monnin et al., 2001; Mottl et al., 1998; Rudnicki et al., 2001; Wheat et al., 2000, 2002, 2003; Wheat and Mottl, 2000).

In addition, fluid samples that were subjected to ¹⁴C analysis demonstrated that, although there is a progression in fluid age from west to east along the western end of the Leg 168 transect, fluids from Site 1030 are younger than those to the west, and fluids from Site 1026 are younger still (Fig. F2C) (Elderfield et al., 1999). It is not possible for waters recharging the basement aquifer near the western end of the Leg 168 transect to gain "youth" as they travel to the east and become increasingly altered; another source of hydrothermal recharge is required. Wheat et al. (2000) showed that there is geochemical evidence for along-strike (south-to-north) fluid transport in basement. Fisher et al. (2003) presented thermal data and calculations based on the hydrogeologic properties of basement rocks and sediment and showed that recharge of Baby Bare (and Site 1026) basement fluids most likely occurs ~50 km to the south, through Grizzly Bare outcrop (Fig. F3A).

Interpretation of fluid ages and rates of fluid flow in basement is difficult on the basis of ¹⁴C data alone because interpretation of radiotracer data from heterogeneous water-rock systems requires application of enormous corrections to account for diffusive and dispersive losses (e.g., Bethke and Johnson, 2002; Fisher, 2004; Fisher et al., 2003; Sanford, 1997; Stein and Fisher, 2003). Actual particle velocities within ridge-flank hydrothermal systems may be 100–10,000 times greater than indicated by simple plug-flow considerations of apparent fluid ages.

Collectively, geochemical data collected along the Leg 168 transect suggest that there are distinct "regions" of hydrothermal circulation within the upper basement and that fluids within each of these regions are hydrogeologically isolated from each other. The first region contains relatively young water at the western end of the drilling transect that has reacted minimally with the formation at 15°–40°C. This fluid becomes older, warmer, and more reacted to the east. The second region is associated with the first buried basement ridge below Sites 1030 and 1031. Fluid from this crustal region reacted with basement at temperatures of 65°–70°C, but the fluid must have cooled during or after ascent from depth because upper basement temperatures are only 40°C. This fluid is young relative to the less reacted fluid to the west. The third chemically distinct region is within crust below Sites 1026 and 1027. This younger fluid mixes vigorously within upper basement at temperatures near 65°C.

Although there was no wireline logging in basement during Leg 168 because penetration was too shallow, borehole packer experiments and analyses of open-hole thermal data were completed to determine local hydrogeologic properties (Becker and Davis, 2003; Becker and Fisher, 2000; Fisher et al., 1997). These experiments indicated near-borehole formation permeabilities of 10^–14 to 10^–10 m², with the highest permeabilities determined for the youngest sites, at the western end of the drilling transect. The data are consistent with the rest of the global seafloor data set and also suggest two notable trends (Fig. F5): a decrease in uppermost basement permeability with increasing age and spatial scaling of permeability estimated using different methods (Becker and Davis, 2003; Fisher, 1998, in press).

Borehole (CORK) observatories were installed during Leg 168 at western Sites 1024 and 1025 and at eastern Sites 1026 and 1027 (Davis and Becker, 2002, 2004). These systems were instrumented to monitor borehole fluid pressure and temperature and to collect long-term fluid samples. Consideration of borehole fluid responses to tidal pressure variations and to regional tectonic events suggests that basement around the boreholes has even higher effective permeability than suggested by packer experiments and thermal logs (Davis et al., 2000, 2001). Because the different estimates of formation permeability were made using different methods and assumptions, it remains unclear if the apparent scaling tells us something important about the nature of basement permeability or is an artifact. This question is being addressed by IODP Expedition 301 and related experiments.

Similarly high permeabilities, on the order of 10^–9 m², were inferred on the basis of steady-state numerical models that used a conductive proxy for coupled heat-fluid flow and extrapolation of relations between permeability and mixing efficiency at lower permeabilities (Davis and Becker, 2002; Davis et al., 1997b; Wang et al., 1997). Fully coupled, transient models of ridge flank circulation have shown that permeabilities this high need not be present throughout the upper crust; if fluid flow is highly channeled (Tsang and Neretnieks, 1998), considerable efficiency in heat transport can be achieved by advection through a small fraction of the crust (Fisher and Becker, 2000; Fisher et al., 1994; Spinelli and Fisher, 2004). In fact, having very high permeability distributed pervasively throughout the upper oceanic crust actually limits the efficiency of lateral fluid flow in transporting heat (Davis et al., 1999; Fisher and Becker, 1995; Rosenberg et al., 2000).

Hole 1026B yielded some of the first direct microbiological observations of ridge-flank fluids. Rock and fluid samples collected during Leg 168 indicated the possible presence of microbes (Bach and Edwards, 2003; Fisk et al., 2000), and a "BioColumn" experiment assessed microbial biomass and diversity in fluids venting from the CORK observatory. Cells collected from the BioColumn included bacteria and archaea, comprising nitrate reducers, thermophilic sulfate reducers, and thermophilic fermentative heterotrophs (Cowen et al., 2003). These tantalizing results encouraged study of the basement biosphere during IODP Expedition 301.