SCIENTIFIC AND OPERATIONAL OBJECTIVES OF EXPEDITION 301

This section highlights the fundamental scientific goals of IODP Expedition 301 and related experiments. Some of the discussion includes operations that are to take place during future drilling and submersible/remotely operated vehicle (ROV) expeditions. This information is included so that the rationale for Expedition 301 operations, and the extent of our successes, will be clear.

Despite extensive survey and drilling work in the eastern flank of JFR prior to Expedition 301, we began the expedition with little information on the geological nature of permeable pathways, the depth extent of fluid circulation, the magnitude of permeability anisotropy, or the significance of hydrogeologic barriers in the crust. We knew that the upper oceanic crust is home to diverse microbiological communities, but we knew little about their populations or ecology or how their distribution relates to primarily crustal stratigraphy, fluid flow paths, water chemistry, or rock alteration. We did not know the concentrations and nature of nonliving organic matter within crustal fluids or how this material influences and is influenced by ocean carbon cycling. We did not understand the scales over which solute transport occurs in oceanic basement rocks or how transport and mixing are influenced by crustal structure and fabric. We did not know how to relate seismic velocities and velocity anisotropy to hydrogeologic properties.

Although some of these topics were addressed during ODP Leg 168 and related surveys and experiments, earlier drilling included penetration of only the uppermost few tens of meters of basement, leaving many questions unresolved. For example, CORK observatories installed at Sites 1026 and 1027 allow determination of upper basement temperatures and fluid pressures, but despite evidence for extremely rapid fluid convection in basement, it was not possible to determine if the dominant fluid flow direction is from Site 1026 to 1027, in the opposite direction, or perhaps perpendicular to both sites. Work during IODP Expedition 301 and related surveys and experiments will help resolve this quandary and will address all of the topics listed above.

Basement work at Second Ridge sites (1026, 1027, and U1301) was given the highest priority and focused on numerous questions, including:

Successfully addressing these and related questions required a combination of conventional and nonstandard scientific drilling approaches. In the rest of this section, we highlight unusual aspects of Expedition 301 and related experiments. Other components of the scientific program that are essential to the overall effort (e.g., petrographic analysis, inorganic geochemistry, and measurements of sediment temperature) are well established within the scientific drilling community and are not discussed. Some of the discussion includes work to be done during future drilling or submersible operations; this information is presented to provide context for planning and prioritization during Expedition 301 and beyond.

Two primary kinds of active hydrologic tests were completed prior to Expedition 301 in the Deep Sea Drilling Project (DSDP) and ODP: slug tests and injection/flow tests (Becker and Davis, 2003; Fisher, 1998). Both kinds of tests have involved a single borehole and a drill string packer. Another form of flow test involves monitoring the movement of fluid into or out of a borehole resulting from natural pressure differences, after removal of a low-permeability sediment seal (e.g., Becker and Davis, 2003; Fisher et al., 1997). During a slug test, formation pressure is abruptly modified through rapid injection of a small fluid volume, and the pressure-time response of the isolated interval allows estimation of transmissivity (T) and storativity (S). Transmissivity is hydraulic conductivity (ease of fluid flow) multiplied by aquifer thickness within a horizontal system. Storativity is a measure of aquifer and fluid compressibility within a horizontal system. Although slug tests can be used to assess transmissivity in the immediate vicinity of a borehole, they are notoriously poor at constraining storativity and are not very useful in formations that are highly permeable. Single hole injection and flow tests provide little information on storage properties, although they can be used to estimate bulk transmissivities with a radial scale that is somewhat greater than slug tests. The radius of investigation of any seafloor hydrologic test depends mainly on T and S and the duration of the test (Becker and Davis, 2003; Fisher, 1998).

CORK observations of formation pressure response to tidal forcing have also been used to estimate hydraulic diffusivity (T/S) and storativity within oceanic crust (e.g., Davis et al., 2000), but as with interpretation of packer experiments, interpretation of passive CORK observations requires assumptions regarding the hydrologic homogeneity and isotropic nature of oceanic crust, as well as the geometry of the flow system and the magnitude, timing, and location of the source function. Observational data (geological, geochemical, and geophysical) demonstrate that the oceanic crust is highly heterogeneous and anisotropic (e.g., Becker and Davis, 2003; Fisher, 2004; Fisher and Becker, 2000). Numerical models have not helped to resolve discrepancies between properties estimated using different techniques because the models themselves are highly idealized. Applying a suite of techniques is the best way to assess the true nature of crustal permeability, including scaling phenomena (Fig. F5), and the validity of simplified representations of these systems.

Two ODP expeditions, Legs 169 and 171A, completed uncontrolled (and largely unplanned) cross-hole experiments using CORK observatories (Fouquet, Zierenberg, Miller, et al., 1998; Moore, Klaus, et al., 1998). On each of these expeditions, previously installed observatories were used to monitor formation fluid pressure tens of meters or more from a site of active drilling. IODP Expedition 301 was the first to be designed with cross-hole experiments being a primary goal. Through use of multilevel CORK observatories, we will isolate and monitor discrete depth intervals in basement, allowing assessment of both vertical and horizontal hydrogeologic connections between sites. Because we anticipated very high basement permeabilities and the perturbation resulting from drilling and observatory installation will take many months to dissipate (Davis and Becker, 2004), the complete cross-hole experiment can not be run until after Expedition 301. This will allow pressure, chemical, and thermal equilibration of the formation below newly installed CORKs so that small changes resulting from active experiments can be readily detected.

During a future drilling expedition, a 24 h packer test will be initiated in a borehole drilled at Site SR-2, between Sites 1026 and U1301 on Second Ridge. This 24 h packer test will be ~50 times longer than any injection test into oceanic crust prior to Expedition 301. Based on a range of apparent bulk properties from packer tests, flow tests, tidal response, and numerical models (Becker and Davis, 2003; Becker and Fisher, 2000; Davis and Becker, 2002; Davis et al., 2000; Fisher et al., 1997), a readily measurable pressure response should be apparent at Sites 1026, 1027, and U1301 (Fig. F6). By subsequently waiting an additional 6–12 months for borehole equilibration, then opening one or more vent valves in a Site SR-2 CORK, we can release overpressured formation fluid for a year or more. This will initiate a free flow ("artesian") well test that will allow an even larger-scale assessment of crustal properties (Figs. F5, F6).

The results of these experiments will have implications well beyond oceanic crust, as there is an ongoing debate concerning the scaling of hydrogeologic properties within heterogeneous systems (Butler and Healy, 1998; Clauser, 1992; Neuman and Di Federico, 2003; Renshaw, 1998; Rovey and Cherkauer, 1995). Results of cross-hole testing in basement holes, combined with other observations and modeling results, can also be used to test equivalent-porous-medium and other representations of the fractured upper crust. The seafloor is an ideal place to address these issues because a single test can be run for a very long time, effectively delineating the scale-dependence of hydrologic properties using a single measurement method. Such tests are generally not possible on land because of logistical and environmental concerns and a lack of demonstrated horizontal continuity. Generating cross-hole data within seafloor boreholes will also allow application of models developed for use in fractured aquifers (Barker, 1988; Moench, 1984) that have not previously been applied to the seafloor.

Based on estimates of prokaryotic biomass in ODP sediment cores (Fig. F7), it has been suggested that the marine subseafloor biosphere is enormous, perhaps exceeding the cumulative biomass of all other ecosystems on Earth (Parkes et al., 1994; Whitman et al., 1998). This hypothesis is compelling, but it remains highly speculative, as it is based on global extrapolations from marine sediments at a relatively small number of sites. Extrapolation of microbiological conditions and densities from the sediment section into basement is also problematic. Cell densities in basement are likely to be low where fluids are old and carbon and nutrient sources are limited, but ridge-flank locations where basement fluids are young and nutrients are abundant might support considerably greater biomass. Also, the size of the subseafloor microbial biosphere does not equate to activity or importance. Cell-count studies in sediments have revealed cell densities exceeding 10⁵/mL, but it is difficult to distinguish living cells from inactive or dead ones. The size of the subseafloor biosphere also tells us little about the magnitude of biogeochemical fluxes into or out of the system.

Little is known about microbial community composition and microbial metabolism in hydrothermally active ridge flanks. The importance of thermal, lithologic, hydrogeological and geochemical controls, and relations between sediment- and basement-hosted communities remains to be determined. Can cell-density trends from marine sediments be extrapolated into basement, or do cell densities increase (or decrease) rapidly once hydrologically active intervals are encountered? The only way to address these questions is through careful sampling and monitoring of sediments, basement rocks, and pore fluids. Five primary techniques are being used to assess the microbiological state of subseafloor environments: (1) total cell counts, (2) cultivation experiments, (3) fluorescence in situ hybridization (FISH), (4) molecular biological techniques based on ribonucleic acid (RNA) and deoxyribonucleic acid (DNA) sequences, and (5) in vitro rate measurements of biological activity.

Recovering pristine formation fluids from oceanic basement is challenging because any formation having significant permeability will be deeply invaded by fluid pumped during drilling, casing, and other operations, including cold, oxygenated seawater and freshwater (the latter often used with drilling mud). Petroleum wells are typically "produced" for weeks or months to recover pristine formation fluids, and in fact, this is the best approach for recovering uncontaminated formation fluid and microbiological samples from a hole in permeable ocean crust. Sites 1026 and U1301 are overpressured; fluids can be extracted for days to months to years from CORKs installed at these sites, minimizing contamination, allowing assessment of population changes over time, and facilitating long-term testing.

Expedition 301 and related experiments include three principal stages of microbiological and biogeochemical study. The first stage comprises biological sampling and analysis of the sediment column, in combination with pore fluid chemistry and in situ temperature measurement, to resolve the nature of sedimentary microbiological communities (e.g., D'Hondt, Jørgensen, Miller, et al., 2003) and compare them to those found in basement. Collection of uncontaminated sediment during drilling requires care but is possible with careful planning (House et al., 2003; Smith et al., 2000). With regard to Expedition 301 objectives, studies of microbial populations in sediment are most compelling in the near-basement environment.

The second stage of microbiological studies comprises sampling and analysis of basement rock. Collection of basalt samples for direct microbiological study is also complicated by contamination, but the use of perfluorocarbon tracer (PFT) during coring operations helps to determine the more and less contaminated samples (Smith et al., 2000). Results will be compared to laboratory and in situ incubation experiments (Edwards et al., 2003, 2002) and to lithologic and alterations studies (e.g., Alt and Teagle, 1999, 2003; Fisk et al., 1998; Furnes and Staudigel, 1999), to relate basement alteration and microbial activity.

The third stage of microbiological studies involves time-series analyses of biological communities and formation fluids. Time-series studies are initiated at the time of CORK installation and will continue for years. CORK sensor strings contain continuous fluid samplers and colonization substrate. These samples will document community succession and biogeochemical response to formation recovery following drilling operations. Additional sampling will occur when we open CORK valves at Site SR-2 6–12 months after installation of these systems. This will allow attachment of sensors and samplers at the seafloor to collect time-series data, fluids, and microbiological materials from multiple depths. After completion of initial long-term flow experiments, additional seafloor valves at other sites can be opened. Through combined molecular biological and geochemical approaches and modeling (Cowen et al., 2003; Reysenbach and Shock, 2002), these samples will provide new insights into the responses of microbial ecosystems to geochemical conditions and the influence of microbial activity on fluid and basement geochemistry. Sampling and analysis of dissolved organic matter, an integrated and long-lasting signature of microbial activity, will elucidate carbon cycling in the crust. In addition to in situ incubation, flow-through incubation at the seafloor will allow system monitoring and manipulation and may allow additional options for temperature and pressure control once power to the site is provided by a (planned) cabled observatory system. This third stage of microbiological sampling and analysis is the only one that will collect high-quality samples from the hydrogeologically most important stratigraphic intervals in the crust.

The extent of water mixing and water-rock interaction within an aquifer depends on properties such as effective porosity (the fraction of open space involved in fluid flow) and dispersivity (mechanical mixing and spreading of water and solutes by diffusion). Understanding these properties is critical to successful reactive-transport modeling and to understanding the age distribution of fluids in the seafloor, but these properties have never been assessed using tests in any DSDP or ODP hole. Effective porosity varies with flow direction in heterogeneous systems as a result of flow channeling (Tsang and Neretnieks, 1998) and must be tested directly. Like permeability, dispersivity varies as a function of test scale and must be determined at the scale of interest (Gelhar et al., 1992; Neuman, 1990; Novakowski, 1992). Tracer experiments will help to quantify rates of fluid transport in basement. We consider tracer tests in a broad sense to include the use of natural tracers, tests initiated through standard IODP operations (e.g., pumping surface seawater during drilling), and experiments involving injection and sampling of specific compounds.

Wheat et al. (2003) used major element chemistry determined on samples recovered by long-term deployments of OsmoSamplers in several sealed ODP Leg 168 boreholes to estimate the rate of equilibration of borehole fluids and flow rates within the surrounding formation. Fluid chemistry changed rapidly during the first 40–60 days after borehole sealing, as drilling fluid was replaced by formation fluid. A slower rate of chemical evolution was documented over the subsequent 1150 days as fluid continued to move through the borehole and borehole fluid in the open hole mixed vertically with fluid in the casing. Formation flow rates estimated from these experiments are comparable to estimates based on other geochemical and thermal considerations (Wheat et al., 2003).

Carbon-14 was used as a natural tracer to estimate rates of transport along the ODP Leg 168 transect (Elderfield et al., 1999), but consideration of nonconservative and mixing behavior must be included in analysis of flow within this heterogeneous crustal system (e.g., Stein and Fisher, 2003). For example, kilometer-scale tracer studies within the Mirror Lake fractured rock system suggest that effective chemical diffusivity of the rock matrix may be much greater than initially thought (Becker and Shapiro, 2000; Shapiro, 2001). Rock transmissivity is highly heterogeneous, and high effective diffusivity may result from preferential tracer migration along a few fractures, the rest being well connected only over short distances (Shapiro and Hsieh, 1998).

We have used (and will use) different tracers in different holes, and at different depth intervals so that single hole, cross-hole, and cross-level transport can be differentiated and quantified. Tracers injected during Expedition 301 included surface seawater, drilling mud, and PFT. Tracers being injected within CORK borehole observatories comprise a mixture of rare earth elements. Candidate tracers for injection at Site SR-2 during a future drilling expedition include one or more varieties of PFT, Br salts, SF₆, ³He, and rhodamine-WT, all of which are readily transported to the ship, environmentally benign, easily introduced into fluids pumped into the borehole, and detectable at low concentrations.

Tracers injected at Site SR-2 might not be sampled at Sites 1026, 1027, or U1301, but we have good reason to be optimistic for success over the long term. Fluid flow velocities in basement along Second Ridge appear to be hundreds to thousands of meters per year (Fisher et al., 2003). Similar rates were inferred in basement along the western end of the Leg 168 transect (Stein and Fisher, 2003). By isolating limited depth intervals in the CORK observatories and instrumenting them with long-term samplers, we optimize chances for detection of cross-hole fluid transport. In addition, each observatory is being used for single-hole tracer experiments (Altman et al., 2002; Novakowski et al., 1998). Interpretation of these kinds of tests, like those for cross-hole tests, is accomplished using forward and inverse modeling techniques, to obtain a match to the observed solute-time history (Becker and Shapiro, 2000; Clemo and Smith, 1997). Future tracer experiments can be run using seafloor pumps once the three-dimensional borehole network is established.

Much of what we know about oceanic crustal stratigraphy is based on seismic refraction and reflection data, but correlations between lithology and physical properties are often ambiguous. Relating outcrop or core properties to seismic-scale measurements is difficult (Jarrard et al., 2003), and relations between seismic and hydrologic properties are essentially unconstrained. There has long been evidence for anisotropy in seismic velocities in the upper crust, with faster velocities in the along-strike direction (e.g., Detrick et al., 1998; Stephen, 1985), but it remains to be determined whether the same crustal fabric includes significant pathways for fluid flow. Expedition 301 included a conventional vertical seismic profile (VSP) experiment to help assess interval velocities and identify gross seismic layering in the upper crust; a future expedition will include an offset-VSP to assess seismic velocity anisotropy. The conventional VSP uses one or more geophones clamped within an open or cased hole and a seismic source at the surface. We used the three-component Well Seismic Tool (WST) and an air gun source run from the drillship. Conventional VSP data from Sites U1301 and SR-2 may allow us to assess earlier interpretations of a seismically distinct boundary at 600 m into basement based on multichannel seismic (MCS) data (e.g., Davis et al., 1996). An offset-VSP, run during the second drilling expedition with the assistance of colleagues on a shooting ship that circles the drillship, will help to assess anisotropy in seismic properties.