Paleomagnetic and rock magnetic measurements at Site 1206, situated south of the summit of Koko Seamount and 6.2 km from DSDP Site 308, focused on assessing the natural remanent magnetization (NRM) of basement rock and obtaining a preliminary paleolatitude estimate for the site during basement formation. Based on nannofossils identified in the washed section of Core 197-1206A-1W and the polarity sequence of minicores sampled, the top of the basement may have formed during magnetic polarity Chron C20n, C21n, or C22n (Berggren et al., 1995; Cande and Kent, 1995) (see "Biostratigraphy"). Priority was given to shore-based sampling for high-resolution paleomagnetic and rock magnetic study. However, a preliminary estimate of paleolatitude for Koko Seamount is presented below based on paleomagnetic analysis of shipboard samples.
Low-field volume-normalized magnetic susceptibility was measured on oriented samples (72 basalt and 23 volcaniclastic samples) after alternating-field (AF) demagnetization with the Kappabridge KLY-2 magnetic susceptibility meter. Three measurements were taken for each sample and then averaged. Measurements were repeatable to within 1% of the average value. The range of magnetic susceptibilities of basalt is 2.0 x 10-3 to 39 x 10-3 SI, with an arithmetic mean of 14.4 x 10-3 SI. The range of volcaniclastic samples is 0.34 x 10-3 to 16 x 10-3 SI (mean = 5.57 x 10-3 SI).
The Koenigsberger ratio (Q) was calculated for each sample using a present geomagnetic field intensity of 31.4 A/m (International Geomagnetic Reference Field) (Barton et al., 1995). Koenigsberger ratios of basalt range from 1.8 to 72 (mean = 13.0). Volcaniclastic samples have Koenigsberger ratios that range from 0.3 to 177 (mean = 19).
Median destructive fields (MDF) range from 7.5 to >80 mT (above the maximum limit of the AF demagnetizer in line with the superconducting quantum interference device [SQUID] magnetometer) (mean = 29.5 mT). The MDF values of volcaniclastic samples range from 9 to 78 mT (mean = 25.9 mT). Some samples from this site show much higher MDF values as compared to samples from previous Leg 197 sites; the reason for this is not clear given the limits of our preliminary data. We also calculated the ratio of the magnetic component demagnetized between 0 and 5 mT to that demagnetized between 20 and 50 mT (
RM) (see "Discussion"). Figure F51 shows a plot of volume magnetic susceptibility (K), the Koenigsberger ratio (Q), MDF, and RM vs. depth. Q-ratios and MDF values, in general, suggest that a majority of the samples carry a stable remanence that can be used to estimate paleolatitude.
Lowrie-Fuller tests (Lowrie and Fuller, 1971) as well as isothermal remanent magnetization acquisition and backfield demagnetization used to measure coercivity of remanence were performed on some samples selected from the lava flow units having a high MDF of NRM (Table T9). In all cases, the MDFs of anhysteretic remanent magnetizations (ARMs) and saturation isothermal remanent magnetization (SIRM) are smaller than the MDFs of NRM.
Some samples (e.g., Samples 197-1206A-3R-2, 99-101 cm, and 16R-5, 75-77 cm) (Figs. F52A, F52D, F53A, F53D) show a rapid decrease of ARM and SIRM and a fairly low coercivity of remanence. This behavior, together with the high MDF of NRM, suggests a mixture of magnetic grains having different sizes. In other samples, the sigmoidal shape of the ARM and SIRM decay and the high value of the coercivity of remanence indicates the presence of single-domain grains (e.g., Samples 197-1206A-4R-5, 55-57 cm; 9R-2, 29-31 cm; 22R-1, 117-119 cm; and 28R-1, 97-99 cm) (Figs. F52B, F52C, F52E, F53B, F53C, F53E). These samples also show little or no decrease in NRM when demagnetized up to 20 mT (Samples 197-1206A-4R-5, 55-57 cm, and 22R-1, 117-119 cm) (Fig. F54), suggesting single-domain behavior.
The behavior of Sample 197-1206A-28R-1, 97-99 cm (Fig. F52F), is more complex. The MDF of NRM is very high. The ARM does not decrease before demagnetization to 25 mT, but the SIRM shows a continuous, slow decrease with AF demagnetization. This may be explained by the presence of a high-coercivity mineral (such as goethite) in the dark brown clay observed in the sample matrix in addition to titanomagnetite (see "Physical Volcanology and Igneous Petrology").
Most of samples have a soft magnetic component that is demagnetized after treatment to 5 mT (Figs. F54, F55, F56, F57). The ratio of this soft component (0-5 mT) to a part of the stable component (20-50 mT) (RM) is useful to illustrate the relative contribution of the soft component to the total NRM (Fig. F55B). RM derived from paleomagnetic data from lava flow and volcaniclastic samples is not strongly correlated to the compressional wave velocity (VP) measured with a PWS3 system (see "Physical Properties"; Fig. F55A). These observations contrast with those from Site 1205 (see "Paleomagnetism and Rock Magnetism" in the "Site 1205" chapter) and attest to the relatively unaltered state of the rocks. The inclination of the 0- to 5-mT soft component from samples with log (RM) > 1.0 tends to be steep and positive (>70°). Magnetic minerals responsible for the origin of the soft components will be studied on shore.
Lava flows and sedimentary units comprising basement (see "Physical Volcanology and Igneous Petrology") were sampled for paleomagnetic analysis. Four or more samples were measured for most lava flows. Fewer samples (one to three) were taken for shipboard analysis from the units dominated by lapilli breccia with basalt lobes and lava flow units that showed extensive fracturing due to the expansion of clay; material was reserved in these units for high-resolution shore-based studies. Sample inclinations for each basement unit were averaged using the method of McFadden and Reid (1982).
Minicore samples of lava flow and volcaniclastic sediment were measured using the 2-G Enterprises SQUID magnetometer. Each minicore was demagnetized by stepwise AF demagnetization following the same measurement routine employed at Site 1205. A total of 72 lava flow and 23 volcaniclastic minicores were measured.
Orthogonal vector diagrams were used to characterize the demagnetization behavior and assess the number of magnetization components represented (Figs. F57, F58). Normal and reversed polarities were recorded. Reversed polarity samples showed a small normal polarity overprint that was demagnetized after 5-10 mT. Normal polarity samples often displayed curved behavior at low-field steps (Fig. F59A, F59B, F59D), suggesting the presence of more than one magnetic component. Primary characteristic remanent magnetization (ChRM) directions were fit using principal component analysis (Kirschvink, 1980) from 35 or 40 mT through 80 mT on many of the normal polarity lava flow samples.
Volcaniclastic samples, in general, had weak NRM values (one to two orders of magnitude less than the NRM values of the lava flows) and were often demagnetized to <10% of the NRM intensity at low-field steps (Fig. F56). Lava flow samples showed uniform demagnetization toward the origin after the removal of a small overprint (Fig. F57). The presence of a high-coercivity mineral is suggested by the inability of AF demagnetization to remove the remanence to <10% of the NRM value in ~60% of the samples (Fig. F58). In three samples (e.g., Fig. F59C), <50% of the total NRM was demagnetized by the last measurement step.
Three polarity intervals were recognized in the Site 1206 basement section. Samples from Cores 197-1206A-2R through 6R are normal polarity. Core 197-1206A-7R through Sample 197-1206A-9R-1, 29-31 cm, are reversed polarity. Sample 197-1206A-9R-2, 110-112 cm, is normal polarity. Because of limited recovery and/or unoriented core pieces in Cores 197-1206A-10R through 14R, samples were not taken for shipboard analysis. Samples from Cores 197-1206A-15R through 43R are normal polarity. Based on the observed polarity sequence and time constraints based on nannofossils from the washed core (see "Biostratigraphy"), the sequence of magnetic chrons recorded may be C20n-C20r-C21n, C21n-C21r-C22n, or C22n-C22r-C23n. Given a radiometric age of 48.1 ± 0.8 Ma based on dredge samples from Koko Seamount (Clague and Dalrymple, 1973), one of the latter two polarity chron assignments is more probable.
ChRM inclinations for volcaniclastic sediments were generally fit between 15-45 or 50 mT. Only samples with maximum angular dispersion angles <10° were considered reliable. In some samples (three of the volcaniclastic samples measured), the demagnetization vectors were stable but did intersect the origin. Nine samples had a stable magnetization with vectors going into the origin. The remaining samples (11) did not have a stable demagnetization.
Sediment generally averages more geomagnetic time than a single lava flow. If we give equal weight to data from each of the volcaniclastic sediment samples that have reliable ChRM inclination fits, the mean inclination is 42.5° (95% confidence interval = ±2.5°) using the method of McFadden and Reid (1982). This suggests a paleolatitude of 24.6° (95% confidence interval = ±1.9°; N = 9) (Table T10).
ChRM inclination directions from lava flow samples were fit between 30 or 35 mT through 80 mT. Inclinations were grouped first according to petrologic units (see "Physical Volcanology and Igneous Petrology") and averaged using the method of McFadden and Reid (1982). Two petrologic units were further broken down in terms of magnetic time units on the basis of polarity change (basement Unit 4) and stratigraphic inclination groupings (basement Unit 6) (see Table T11). Fourteen magnetic time units were identified, with a mean inclination value of 38.5° (95% confidence interval = +8.4°/-10.9°), suggesting a paleolatitude of 21.7° (95% confidence interval = +6.4°/-7.0°) (Table T11).
Paleolatitudes based on sediment may suffer from a bias toward shallower values because of detrital inclination error (King, 1955) or compaction-induced inclination shallowing (Anson and Kodama, 1987). The paleolatitude of the volcaniclastic interbeds (24.6°; 95% confidence interval = ±1.9°), however, is similar to that based on the basalt units (21.7°; 95% confidence interval = +6.4°/-7.0°). But because of the small number of samples used in the inclination average, the uncertainty of the paleolatitude based on the volcaniclastic units may be underestimated.
The angular polar dispersion (S) of the paleomagnetic data from the lava flow units can be estimated by transforming the estimate of the directional precision parameter (k) (McFadden and Reid, 1982) into pole space (Cox, 1970; Tarduno and Sager, 1995). For 14 inclination units, S = 16.2°. This value is slightly higher than that expected based on global data from 45- to 80-Ma lava flows (S = 12.2° ± 1.6°) (McFadden et al., 1991). The distribution of lava flow inclinations appears to represent well the range of values expected from a Fisher distribution (Fisher, 1953) (Fig. F60). Some serial correlation may exist between some of the lava flow units sampled (e.g., Units 5 and 6, 10, and 11). However, the limited number of samples collected per lava flow unit for shipboard analysis limits further consideration of this aspect; this will be addressed in shore-based studies.
The preliminary paleolatitude mean derived from the Site 1206 lava flow samples is ~3° from the present-day latitude of the Hawaiian hotspot (~19°). Paleomagnetic data from Sites 1203, 1204, 1205, and 1206 are consistent with Late Cretaceous-early Tertiary hotspot motion during the formation of the Emperor Seamounts (Tarduno and Cottrell, 1997; Cottrell and Tarduno, in press). The preliminary paleomagnetic data from Site 1206 hint at a slowing of hotspot motion at 48 Ma, near the bend in the Hawaiian-Emperor chain. This possibility will be further examined in shore-based studies utilizing thermal demagnetization.
