RESULTS

The PFT concentrations measured for all sediment samples are listed in Table T1, and a summary of the data is also shown in Table T2, sorted by sampling location (center of the core vs. core periphery), drilling technique (advanced piston corer [APC] vs. extended core barrel [XCB]), and site (deepwater sites vs. continental margin and shelf sites). This comparison reveals three factors that influence the potential for seawater contamination of sediment cores.

1. The center of the sediment cores contain less PFT than the core margins and have lower potential seawater contamination. The difference can range from a factor of 3.5 (Site 1231; APC cores) to almost 100 (Site 1229; APC cores). This result appears to be independent of the drilling technique and applies to APC as well as XCB cores.
2. Samples obtained by APC coring are generally less contaminated than sediments obtained by XCB coring from the same site. For core center samples, the difference was approximately a factor of 10 at Site 1226 and a factor of 3 at Site 1230 (Table T2). XCB coring is more disruptive and introduces more seawater into the sediment core.
3. The gradient between average PFT concentrations in the core center and the periphery is steeper in sediment cores from the Peru Margin sites on the continental shelf (a factor of 65) than in cores from deepwater sites (a factor of 9) (Table T2). This difference appears to be related to water depth and therefore to core retrieval time; whereas Peru Margin Sites 1227, 1228, and 1229 were in 426, 252, and 150 m water depth, the deepwater Pacific Sites 1226, 1230, and 1231 had water depths of 3296, 5071, and 4827 m. In sediment samples from deepwater sites, the PFT gradients appear to dissipate during core retrieval (1-2 hr), compared to fresher and steeper PFT gradients of Peru Margin sediment samples (<30 min retrieval time).

The potential seawater contamination for Leg 201 sites is in a similar range as previous data for ODP Site 1149 southeast of Japan on the seaward side of the Izu-Bonin Trench at a water depth of 5830 m (Table T3) (Smith et al., 2000b).

For six cores from deepwater Holes 1226E and 1230B, PFT samples were obtained with 1-mL syringes diagonally across freshly cut core sections on the catwalk. After sampling, the sediment samples were placed into 2-mL GC vials, closed in the open air to avoid contamination with ubiquitous laboratory PFT, and weighed before GC analysis. The PFT concentrations in core cross sections do not always follow a predictable gradient, with maxima at the core liner and minima in the center of the core (Table T4). On average, the six transects show lower PFT concentrations in the samples from the inner core (samples 2-6; average = 0.012-0.044 ng PFT/g sediment) and elevated PFT counts in the outer samples at the core liner (samples 1 and 7; average = 0.797 and 0.031 ng PFT/g sediment).

Contamination tests are especially significant in evaluating contamination risks for samples that are used for microbial cultivations. Sediment samples are mixed anaerobically (under nitrogen) with sterile saline solution on a 1:4 volume basis. These slurries are then used for a variety of microbial enrichments and quantifications, including most probable number dilutions on defined media or extracts of natural substrates (see "Microbiology" and media recipes in Tables T4, T5, and T6, all in the "Explanatory Notes" chapter).

Table T5 shows the data for potential seawater contamination and microsphere counts for slurry samples and allows an assessment as to whether detectable concentrations of the easily diffusible PFT tracer coincide with visible microspheres. Whereas the chemical tracer indicates the extent of seawater contamination and influx of dissolved compounds, the prokaryote-sized microspheres demonstrate that prokaryotes could actually penetrate the sediment. In other words, they outline the prokaryote contamination potential. Three slurry samples with high potential seawater contamination, in the range of 0.05-0.4 µL seawater/mL slurry (Sections 201-1226B-2H-3 [center]; 201-1229A-11H-5 [center]; and 201-1230A-38X-1 [center]), also contained microspheres, bacterial mimics. These slurry samples are most likely contaminated. In slurry samples with potential seawater contamination levels below ~0.05 µL/mL, microspheres were not found at all or were observed a single time. A possible source of experimental error was found in Section 201-1226B-47X-2, which had high microsphere counts but no detectable PFT. Prolonged gassing of the anaerobic slurry with nitrogen could have removed the PFT tracer, while retaining the microspheres.