RESULTS AND DISCUSSION

Event Study Samples

The AMS data for the event studies are presented in Table T3. Only one of the 61 samples passes the filter, and there are insufficient acceptable data to characterize the events on the basis of magnetic fabric. The event study samples were analyzed on the Roly-Poly, which has a lower sensitivity than the Kappabridge, but many of the samples also come from shallower sites with high carbonate content and therefore poor susceptibility.

Site 1061

Filtered L data show no trend between 104.5 and 121.5 meters composite depth (mcd) (Table T4; Fig. F1). Low K₁ inclination values show that the lineation is nearly horizontal. Filtered F values decrease from 104 to 115 mcd, and then increase noticeably to the base of the interval at 121.5 mcd. Foliation behaves independently of L but similar to that of K_mean. Major inflections in K_mean (e.g., at 115 mcd) are probably largely a factor of increased dilution of terrigenous material by carbonate. Foliation is less likely to be influenced by carbonate content unless carbonate is affecting the compaction properties of the sediment. The similarity between K_mean and F may instead be a function of changes in source material as sediment provenance changed slightly with sea level through the glacial and interglacial intervals.

Normalized (see "Methods") declination values generally do not show a preferred orientation, except for the cluster in the -30° to 0° direction at 115-118 mcd suggesting alignment of most particles in a north-south to north-northwest-south-southeast direction. This differs slightly from the northwest-southeast alignment of the main axis of the BOR that corresponds to the southeastward WBUC flow on the northeastern flank of the ridge. However, Site 1061 was cored on the crest of the BOR where the WBUC crosses to eventually flow northwestward on the southwestern flank. The core may be located where the current is flowing southward as it turns direction, or flow direction may be modified slightly by local topography.

Site 1057

The lineation parameter at Site 1057 exhibits a fair degree of short-term variability, with values generally between 1.003 and 1.013, but there are no general trends in L with depth in the hole (Table T5; Fig. F2). K₁ inclination values indicate that the lineation is nearly horizontal, as is expected in sedimentary environments. Sedimentation rates at Site 1057 are slower than at Site 1061 (Keigwin, Rio, Acton, et al., 1998a), and it is possible that the depositional fabric may have been disrupted by bioturbation.

Unlike at Site 1061, filtered F values at Site 1057 do not indicate any correspondence between F and K_mean. F at Site 1057 generally varies between 1.005 and 1.015, compared with the trend from 1.01 to 11.1 at Site 1061. The smaller F values could in part be due to the shallower sediment depth and lesser degree of compaction of the Site 1057 sample interval (32.5-47 mcd) compared with the interval from Site 1061 deeper in the sediment column (114.5-121.5 mcd). They could also result from a different source of minerals supplied to Site 1057 on the upper part of the rise compared to Site 1061 deeper on the rise.

Normalized declination values appear to follow trends downhole. From 32 to 37 mcd K₁ is aligned northeast-southwest. The alignment disappears from 37 to 39 mcd, but from 39 to 40 mcd there is a trend that starts in a north-northwest-south-southeast direction at 39 mcd and rotates counterclockwise to end in a east-west direction at 47 mcd. The present-day contour lines run northeast-southwest at this site, which is in agreement with contour current parallel alignment at the top of the interval but not at the bottom. Downslope transportation might be expected to align particles at right angles to the contours, but there is no evidence for a large event in this interval.

General

The weak lineation of the sediments may be in part a result of the depositional environment on the crest of the ridge. Although the WBUC is strong enough to winnow sediments on the flanks of the ridge, it may cross the ridge in a more diffuse form rather than as a discrete jet (Haskell and Johnson, 1993). The fact that sediment is being deposited on the ridge crest rather than eroded as observed on the flanks suggests that circulation is not as intense on the crest where the cores were taken, and flow may not be sufficiently intense to align particles.

Sediment composition may also play a role in producing poorly defined lineation records. The work of Johnson et al. (1988) and Haskell et al. (1991) indicates that circulation is most intense during interglacial periods, but sediments deposited during those times have the highest carbonate content, which yields the poorest susceptibility signals. Intervals with better circulation potential for particle alignment may therefore be mineralogically unsuitable for recording particle alignment by magnetic means.

Kappabridge measurement errors are generally within ±6.0 10^-6 SI units. Although the filtering procedure used in this study is not a rigorous statistical test, data points that pass the filter are probably significant. However, the measurement error combined with the relatively small differences between K₁ and K₂ suggests most of the lineation variability probably lies within the error limits. This is also the case for some of the variation in foliation, especially where foliation is weak, but larger variations such as those observed at Site 1061 are probably significant.