STRONTIUM ISOTOPE CHRONOSTRATIGRAPHY

Sixty-eight Sr isotope age estimates were obtained from mollusk shells (~4-6 mg) from the Bethany Beach borehole (Table T3; Fig. F10). Shells were cleaned ultrasonically, powdered, and dissolved in 1.5-N HCl. Samples were centrifuged and introduced into ion-exchange columns. Standard ion-exchange techniques were used to separate the strontium (Hart and Brooks, 1974), and samples were analyzed on a VG Sector mass spectrometer at Rutgers University. NBS 987 is measured on the Rutgers Sector as 0.710255 (2- standard deviation [SD] = 0.000008; n = 22) normalized to ⁸⁶Sr/⁸⁸Sr = 0.1194. Internal precision on the sector for the data set averaged 0.000008; external precision for the Sector has been reported as approximately ±0.000020 (Oslick et al., 1994). Though most of the Sr isotopic analyses yield monotonically increasing values upsection (decreasing age) (Fig. F10), there are several age inversions, even in the lower lower middle Miocene section; the latter is unexpected because the rate of Sr isotopic increase was high in the early early middle Miocene and previous data sets (e.g., Cape May) (Miller et al., 1996a) generally show monotonically increasing values upsection. In addition, at least seven data points are interpreted as statistical outliers (green dots on Fig. F10). The cause of this may be partly due to stratigraphic reworking (e.g., 607, 622.3, and 710.5 ft; 185.01, 189.68, and 216.56 m) but also may be partially attributed to minor alteration of some of the shells (e.g., in indurated zones at 572.8 and 574 ft; 174.59 and 174.96 m). Additional Sr isotopic studies of foraminifers are warranted to test the fidelity of the measurement made on shell material, particularly for the middle-upper Miocene section.

Ages were assigned using the Berggren et al. (1995) timescale (Table T3); the Miocene regressions of Oslick et al. (1994) were used where possible. The Oslick et al. (1994) regressions are only valid to sections older than 9.9 Ma (Sr isotopic values = <0.708930). For younger Miocene sediments, we used a modification of the regressions of Martin et al. (1999). Martin et al. (1999) noted a uniform offset between their middle-late Miocene regression and that of Oslick et al. (1994) of ~0.000050 after correcting for differences between standards (fig. 7 in Martin et al., 1999). This difference could be attributed to interlaboratory calibration problems (i.e., a 0.000020 difference has been reported between University of Florida and other laboratories after correcting for differences in the NBS987 standard; see Oslick et al. [1994] and Howarth and MacArthur [1997] for discussion). Thus, we subtracted 0.000050 from our data and applied the Martin et al. (1999) regressions for isotopic values 0.709050 and 0.708930 (i.e., the Miocene section younger than 9.2 Ma). We note that following this procedure (i.e., adding 0.00050 to our data and using the Martin et al. [1999] regression) for values between 0.708930 and 0.708850 (i.e., the Miocene section older than 9.2 Ma) yields very similar ages to those obtained using the uncorrected data and the Oslick et al. (1994) regression. The sole Oligocene age was estimated using the age regression of Reilly et al. (in press). For the sole Pleistocene analysis (24.6 ft; 7.50 m), we derived a linear regression using the data of Farrell et al. (1995), correcting their data to NBS987 of 0.710255 and fitting linear segments to the data between 0 and 2.5 Ma:

Age = 15235.08636 - 21482.27712 x (⁸⁶Sr/⁸⁷Sr).

Miller et al. (1991) and Oslick et al. (1994) estimate age errors derived from linear regressions of Sr isotopic records. Age errors for 15.5-22.8 Ma are ±0.61 m.y. and for 9.7-15.5 Ma are ±1.17 m.y. at the 95% confidence interval for a single analysis. Increasing the number of analyses at a given level improves the age estimate (±0.40 and ±0.76 m.y. for three analyses each in the two intervals) (Oslick et al., 1994). The regression for the later Oligocene (23.8-27.5 Ma) has an age error ±1 m.y. (for one analysis at the 95% confidence interval) to ±0.6 m.y. (for three analyses at the 95% confidence interval) (Reilly et al., in press). The regression for the late Pliocene-Pleistocene (0-2.5 Ma) has an age error of ±0.35 m.y. (for one analysis at the 95% confidence interval) to ±0.2 m.y. (for three analyses at the 95% confidence interval) (Miller, unpubl. analysis of the data of Farrell et al., 1995).

A strontium value of 0.709147 obtained from a shell at 24.6 ft (7.50 m) corresponds to an isotopic age of 1 ± 0.35 Ma (Fig. F10). This appears to contradict correlations, that suggest a younger Pleistocene age for this section (see "Omar Formation").

The poorly fossiliferous nature of the upper 375 ft (114.3 m) of the borehole otherwise precludes Sr isotopic study. From 375 to 1400 ft (114.3 to 426.72 m), the core contained ample calcareous material for Sr isotopic analysis. Sr isotopes allowed us to provide age estimates for 11-12 Miocene sequences. Samples taken above the sequence boundary at 698.5 ft (212.90 m) have age estimates that vary more than the section below the sequence boundary at 698.5 ft (212.90 m). Age inversions are common above this level, in part reflecting the lower global rate of Sr isotopic change in the younger section.

Sr isotopic age estimates for a sequence from 374 to 452.45 ft (114.0 to 137.91 m; a sequence comprising the lower Manokin formation) range from 9.6 to 11.7 Ma, though they cluster from 9.6 to 10.5 Ma (the 11.7-Ma estimate is an outlier) (Fig. F10). A sequence between 452.45 and 523.05 ft (137.91 and 159.43 m; upper St. Marys Formation) has Sr ages of 9.6 and 10.8 Ma. A linear regression through the data in this and the overlying sequence yields our best age estimate of 9.8-10.2 Ma for the 375- to 452.45-ft (114.3-137.91 m) sequence and 10.2-10.6 Ma for the 452.45- to 523.05-ft (137.91-159.43 m) sequence (Figs. F10, F11) and mean sedimentation rate of ~56 m/m.y. The presence of Globorotalia pseudomiocenica at 506 ft (154.23 m) (FO = 8.3 Ma?) may indicate that the lower sequence is slightly younger than indicated by Sr, though this is based on a sole specimen of uncertain taxonomic assignment.

The sequence from 523.05 to 575.2 ft (159.43 to 175.32 m; lower St. Marys Formation) has wide-ranging ages of 11.7-15.0 Ma. We believe the older ages represent either diagenesis or reworking of older material (Fig. F10). The samples at 541.25 and 567.5 ft (164.97 and 172.97 m) with age estimates of 11.8 and 11.7 Ma are fine-grained silts and clays, making diagenesis or reworking less likely, and we prefer an age assignment of ~11.8 Ma for this sequence. Assuming similar sedimentation rates as found in the two sequences above (56 m/m.y.) yields our best age estimate of 11.6-11.9 Ma for this sequence (Figs. F10, F11), though we admit that the age of this sequence is still poorly constrained. Nevertheless, this sequence is clearly late middle Miocene as supported by the identification of dinocyst Zone DN7 (<12.4 Ma) (see "Biostratigraphy").

The sequence found from 575.2 to 649 ft (175.32 to 197.82 m; uppermost Choptank Formation) yields ages ranging from 13.1 to 13.7 Ma, except for two analyses of 16.4 and 16.5 Ma at 607 and 622.3 ft (185.01 and 189.68 m), respectively. We consider these two analyses to be from reworked or diagenetically altered shells (Fig. F10). A linear regression through the four reliable points yields an age for the best estimate of this sequence of 13.1-13.5 Ma (Figs. F10, F11) and a sedimentation rate of 56 m/m.y., remarkably similar to the sedimentation rates above. This sequence appears to correlate with the Kw3 sequence in New Jersey (Sugarman et al., 1993; Miller et al., 1997).

The sequence from 649 to 698.5 ft (197.82 to 212.90 m; middle Choptank Formation) has Sr ages ranging from 13.4 to 14.7 Ma. The ages are actually inverted in this sequence, although these inversions are within the external precision (error bars, Fig. F10). Assuming similar sedimentation rates as found in the sequence above (56 m/m.y.) yields our best age estimate of 14.2-14.5 Ma for this sequence (Figs. F10, F11), though the age of this sequence is least well constrained of the Miocene sequences at Bethany Beach (i.e., this sequence could be anywhere in the interval 13.4-14.7 Ma). This sequence appears to correlate with the Kw2c sequence in New Jersey (Miller et al., 1997) (Figs. F10, F11; Table T6).

Sr isotopic ages from the section below 698.5 ft (212.90 m) constrain the ages well. A sequence from 698.5 to 787.1 ft (212.90 to 239.91 m; lower Choptank Formation) has Sr ages ranging from 15.8 to 16.2 Ma. A sample at 710.5 ft (216.45 m; 14.3 Ma) is believed to be burrowed down. This sequence appears to correlate with the Kw2b sequence in New Jersey (Miller et al., 1997) (Figs. F10, F11; Table T6).

The sequence from 787.1 to 897.7 ft (239.91 to 273.62 m; lowermost Choptank-uppermost Calvert Formation) ranges in age from 16.5 to 17.3 Ma. Several ages in this sequence are inverted, possibly indicating reworking of younger shells within the sequence. This sequence appears to correlate with the Kw2a sequence in New Jersey (Miller et al., 1997) (Figs. F10, F11). There is no discernable hiatus between this and the overlying sequence. Assuming constant sedimentation rates through both sequences (38 m/m.y.) yields our best age estimate of 15.7-16.4 and 16.4-17.3 Ma for the 698.5- to 787.1- and 787.1- to 897.7-ft (212.75-239.91 and 239.91-273.62 m) sequences, respectively (Fig. F10; Table T6). Alternatively, assuming best fits to the Sr ages within each sequences yields age estimate of 15.8-16.2 Ma (68 m/m.y. sedimentation rate) and 16.7-17.0 Ma (96 m/m.y. sedimentation rate), respectively (dashed lines, Fig. F10; Table T6).

The section from 897.7 to 1153 ft (273.62 to 351.43 m; Calvert Formation, partim) contains three sequences, but the hiatuses between them are too short to resolve using Sr isotopic stratigraphy. The mean sedimentation rate for these three sequences is 57 m/m.y., again remarkably similar to those above. The sequence from 897.7 to 981.3 ft (273.62 to 299.10 m) yields Sr ages between 17.1 and 18.8 Ma. Assuming constant sedimentation rates for the entire interval from 897.7 to 1153 ft (273.62 to 351.43 m) yields a best age estimate of 18.0 to 18.4 Ma for the 897.7- to 981.3-ft (273.62-299.10 m) sequence. There does not appear to be a sequence of equivalent age in New Jersey. The sequence from 981.3 to 1057.95 ft (299.10 to 322.46 m) has Sr ages between 18.6 and 19.2 Ma. Our best age estimate assuming constant sedimentation rates is 18.4-18.8 Ma. This sequence is equivalent in age to sequence Kw1c in New Jersey (Miller et al., 1997). The sequence from 1057.95 to 1153 ft (322.46 to 351.43 m) has ages ranging between 18.4 and 19.4 Ma (note that the age range is between 18.9 and 19.2 Ma if duplicates are averaged). Our best age estimate assuming constant sedimentation rates is 18.8-19.3 Ma. The estimated age of 19.3 Ma for the base of Zone NN3 at 1091 ft (332.54 m) is remarkably consistent with the sedimentation rate model derived from Sr isotopes (Fig. F10). The presence of G. praescitula at 1103 ft (336.19 m) (FO = ~18.5 Ma [Berggren et al., 1995]) does not appear to be consistent with the age model. One possible explanation is that forms assigned to G. cf. praescitula appeared in strata older than 19 Ma at Site 608 (Miller et al., 1991). This sequence from 1057.95 to 1153 ft (322.46 to 351.43 m) is equivalent in age either to sequence Kw1b or Kw1c in New Jersey (Miller et al., 1997); strict interpretation of the ages suggests that it is equivalent to the Kw1c and that the Kw1b is cut out in Delaware (Fig. F11; Table T6).

The sequence(s) between 1153 and 1421.1 ft (351.43 and 433.15 m) range in age from 20.3 Ma at the top to 21.0 Ma at the base. Fitting a linear regression to the data yields a best age estimate of 20.2-20.8 Ma. The sedimentation rate in this sequence is very high (136 m/m.y.). This sequence correlates to the sequence Kw1a in New Jersey (Miller et al., 1997).

Only one age was obtained on the three sequences between 1421.1 and 1465.7 ft (433.15 and 466.75 m). There are no Sr isotopic ages in the upper or lower sequences (1421.1-1430.5 and 1454.5-1465.7 ft; 433.15-436.02 and 443.33-466.75 m). A sample at 1430.8 ft (436.02 m) yields an age of 21.0 Ma, perhaps indicating that this sequence is equivalent to the Kw1a1 subsequence, the lowermost of three subsequences that comprise the Kw1a sequence in New Jersey (Miller et al., Chap. 2, this volume). This suggests correlation of the 1421.1- to 1430.5-ft (433.15-436.02 m) sequence to the Kw1a2 and the 1153- to 1421.1-ft (351.52-433.15 m) sequence to the Kw1a3 sequence. No Sr results from the Bethany Beach borehole yielded ages equivalent to the Kw0 sequence in New Jersey (~22-23.5 Ma) (Miller et al., 1998a). However, the HO of G. kugleri at 1446.6 ft (430.07 m) suggests that these strata are 21.5-23.8 Ma and thus equivalent to Kw0, not Kw1a1.

A Sr age of 29.0 Ma was obtained from a shell at 1467.8 ft (447.39 m), indicating that the base of the borehole is in the Oligocene and that there is a substantial hiatus associated with the Oligocene/Miocene boundary at Bethany Beach. Planktonic foraminifers at 1467.2 ft (447.20 m) are assigned to Zone P22 (23.8-27.1 Ma); the one Sr isotope analysis adjacent to this level is not consistent with this identification but suggests correlation to Biochron P21b.

Sedimentation rates are high in the Miocene at Bethany Beach compared to the coeval sections in New Jersey. Sedimentation rates were 37-59 m/m.y. (mean = 53 m/m.y.) in Delaware from 9.8 to 18.8 Ma and 136 m/m.y. from 20.2 to 20.8 Ma. In contrast, sedimentation rates at the thickest Miocene section in New Jersey, Cape May, were 29-47 m/m.y. (mean = 40 m/m.y.) from 11.5 to 20.2 Ma and 91 m/m.y. from 20.2 to 20.6 Ma. Nevertheless, thickness does not scale into stratigraphic continuity; the New Jersey record is much more complete in the early early Miocene (19-23.8 Ma) with the Kw1b and Kw0 sequence apparently missing in Delaware, though the Delaware section is more complete in the late early Miocene (~19-16.2 Ma) (Fig. F11).