Biotites were hand picked from 27 tephra layers in Leg 165 cores for 40Ar/39Ar analysis. Up to 24 single crystal dates were obtained from each sample. The results of biotite analyses are presented in Table 1, where the weighted mean (weighted on the basis of the analytical error for each analysis) of the single crystal dates is adopted as the radiometric age of each tephra layer. Weighted means and errors were calculated on the basis of the method presented in Ludwig (1998).
The results of 40Ar/39Ar dating of Caribbean tephra layers are compared in Table 2 with the ages assigned to the sediments containing the tephra, on the basis of nannofossil and planktonic foraminiferal biochronology documented by the Leg 165 scientific party and subsequent shore-based studies (Sigurdsson, Leckie, Acton, et al., 1997; Chaisson and D'Hondt, Chap. 2, this volume; Kameo and Bralower, Chap. 1, this volume). In general, each of the 27 dated tephra layers can be assigned to a specific planktonic foraminifer and nannofossil biozone, providing an age range within the adopted biochronology. The new radiometric ages display three types of relationships with respect to the biostratigraphic age ranges. In the majority of the analyzed tephra layers (15 examples), the weighted mean 40Ar/39Ar age lies within the narrowly defined biozone age range for each layer, as shown in Table 2. In most of the other samples the radiometric average age lies slightly outside the narrowly defined biozone age range (Table 2), even when the standard error is taken into account. In several cases, the weighted means are based on a large number of single-crystal analyses and with small error (e.g., Samples 165-999A-48X-6, 42-43 cm; 165-1000B-14R-2, 139-140 cm; 165-1000B-19R-2, 99-101 cm; and 165-998A-27X-1, 120-122 cm). In the majority of cases, these radiometric ages lie outside the designated age range of the planktonic foraminifer age estimates, but within the nannofossil age estimates. Further work is required to establish how close these tephra layers are to biozone boundaries, in order to utilize them to further constrain or tune the biochronostratigraphic time scale.
Biotite ages in three of the tephra layers are much older than indicated by the biochronology and the difference—tens of millions of years—is much greater than can be reasonably accounted for by errors in the current time scale (e.g., tephra layers in Samples 165-998A-13H-3, 32-33 cm; 165-1001A-36R-3, 70-71 cm; and 165-1001B-6R-3, 130-131 cm). There are several reasons why the 40Ar/39Ar ages might be much older. First, altered biotites tend to suffer from recoil more than less altered ones, and recoil loss of 39Ar causes anomalously high 40Ar*/39Ar ratios and thus should produce higher apparent ages. All studies on altered biotites to date show, however, that alteration tends to decrease the apparent age (Adams and Kelley, 1997). Second, biotites may contain excess argon. This may include biotite grains that were in the magma chamber and experienced a high partial pressure of argon or took on mantle argon, but never fully degassed during the eruption. Third, the biotites may be xenocrysts, plucked from older wall rocks or conduit walls during or shortly before eruption and dispersed and deposited with the juvenile tephra. Finally, some of these high-age biotites may be in tephra layers that represent turbidites or other type of sediment from a terrigenous source, and thus not derived directly from a contemporaneous magmatic source.
Feldspar crystals were also separated from 16 tephra layers for 40Ar/39Ar dating. Upon analysis of these grains, it became clear that the majority were sodic plagioclase and thus with a potassium content too low for yielding meaningful dates. One layer contained common sanidine crystals, however, which provided 16 accurate single-crystal dates (Sample 165-999A-48X-6, 42-43 cm), with a weighted mean of 16.36 ± 0.31 Ma for the layer (Table 3). In comparison, the weighted mean age of six biotite crystals from the upper part of this layer (interval 165-999A-48X-6, 42-43 cm) is 17.98 ± 0.50 Ma. The difference in biotite and sanidine ages is significant, and may reflect slight alteration of the biotites.