It is now over twenty years since it was first demonstrated that orbital cycles are preserved in the climatic records of deep-sea sediments [ Shackleton and Opdyke, 1973; Hays et al., 1976]. Hays et al. [1976] adjusted their initial timescale to bring the peak variance of obliquity cycles to the same frequency calculated for obliquity by astronomers, and were thus the first to use cyclostratigraphy as a means of tuning timescales. During the 1980s, astrochronology was largely restricted to the Brunhes Chron and was utilized successfully to constrain the age of the oxygen isotope record with precision of a few thousand years [e.g. Martinson et al., 1987]. The turning point for astrochronology/cyclostratigraphy came with DSDP/ODP hydraulic piston core recovery at multiple holes for single sites with adequate sedimentation rates, oxygen isotopic records and magnetostratigraphies. The hydraulic piston corer allowed good undisturbed sediment recovery, and the drilling of multiple holes allowed complete recovery of composite sections. Ruddiman et al. [1989] and Raymo et al. [1989] compiled oxygen isotope data for the Pleistocene and Upper Pliocene at Site 607 (North Atlantic). Orbital tuning using the strong obliquity signal for the Matuyama Chron did not reveal any significant discrepancy with the standard [ Mankinen and Dalrymple, 1979] K-Ar polarity chron ages, apart from a significantly shorter duration for the Olduvai subchron.
Although the astrochronology of Johnson [1982] had indicated an older age for the Brunhes/Matuyama boundary, the realization that the K-Ar ages compiled by Mankinen and Dalrymple [1979] for Plio-Pleistocene polarity chrons were in error came with the study of Site 677 in the eastern equatorial Pacific [ Shackleton et al., 1990]. At this site, the oxygen isotopic record from benthic foraminifera is dominated by the obliquity signal, and the planktic record by precession. Because the effect of precession is modulated by the eccentricity cycle, the precession-dominated planktic record provided a better means of correlation to the astronomical data than the obliquity record. The Site 677 oxygen isotope data could therefore be matched to the astronomical data with more confidence than the poorly modulated obliquity signal at Site 607. Indeed at Site 607, Ruddiman et al. [1989] apparently miscounted the number of obliquity cycles in the lower Brunhes and early Pleistocene partly because they accepted the K-Ar age for the Brunhes/Matuyama boundary. The Site 677 astrochronology [ Shackleton et al., 1990] yielded an age for the Brunhes/Matuyama boundary (780 ka) at odds with the Mankinen and Dalrymple [1979] K-Ar age (730 ka) (Table 1).
The older astrochronological age for the Brunhes/Matuyama
boundary from Site 677 coincided with the observation of
discrepancies between conventional K-Ar ages and Pliocene carbonate
cyclostratigraphy in the southern Italian Trubi Formation [
Hilgen and Langereis, 1989]. Spectral analysis of CaCO
content
in these Pliocene marly limestones yielded periodicities of 15.5,
18.5, 35.0 and 335 kyr, indicating a consistent discrepancy with
the astronomical solutions of 19, 23, 41 and 413 kyr. Hilgen
and Langereis [1989] computed the age of the Gilbert and Gauss
reversals using an age of 3.40 Ma for the Gauss/Matuyama boundary
and tuning the CaCO
cycles to the astronomical solution.
The Late Pliocene and Early Pleistocene of southern Italy is characterized by sapropels, dark organic-rich laminated layers. Hilgen [1991a] showed that for Late Pleistocene sapropels from the eastern Mediterranean (core RC 9-181), individual sapropels correlate to minimum peak values of the precession index, and small and large scale sapropel clusters correlate to eccentricity maxima related to the 100 kyr and 400 kyr eccentricity cycles. Hilgen [1991a] applied this observation to the sapropel-bearing Late Pliocene-Early Pleistocene sections in southern Italy which had magnetostratigraphic control. The sapropel occurrences and sapropel clusters could be matched to the astronomical solution leading to a new timescale for the 1.8-3.2 Ma interval (Table 1). Hilgen [1991a] showed that the timescales of Berggren et al. [1985] and Raymo et al. [1989] lead to a mismatch of the sapropel occurrences with the astronomical solution.
Below the main interval of sapropel occurrence, Hilgen
[1991b] used the CaCO
cycles in the same southern Italian
sections to generate an astrochronology for the Pliocene (Table 1).
Hilgen [1991b] used the relationship between the sapropel
occurrences and CaCO
content in the Upper Pliocene, to infer
that: (1) the grey marl beds denoting the small-scale CaCO
minima correspond to minima in the precession index and (2) the
larger scale CaCO
minima correspond to maximum amplitude
variations of the precession index related to the eccentricity
modulation. The recognition of both a precession and an
eccentricity signal in the CaCO
record allowed the the
observations to be tuned to the astronomical solution with
considerably more confidence than would have been the case in the
absence of the eccentricity modulation. This important paper
resulted in astrochronological ages for Gauss and Gilbert reversals
(Table 1).
Neogene astrochronology has been extended into the Late
Miocene using cores collected during ODP Leg 138 [ Shackleton
et al., 1994]. Multiple cores at each site were correlated using
GRAPE (Gamma Ray Attenuation Porosity Evaluator) data, magnetic
susceptibility, and color reflectance to produce composite sections
[ Hagelberg et al., 1992]. The composite GRAPE records (which
are controlled by the ratio of calcite to biogenic opal) were then
matched with the orbital insolation solution of Berger and
Loutre [1991]. This matching led to astrochronological age
estimates for geomagnetic reversals in the 3-6.2 Ma (top Kaena to
C3A.n2)
interval (Table 1).
Consistent estimates were determined from several holes/sites, and
estimates are within a few tens of thousands of years of the
estimates given by Hilgen [1991a,b] from southern Italian
sections, and differ from those given by Cande and Kent
[1992a] (Table 1). Shackleton et al., [1994] generated a
seafloor anomaly timescale beyond C3A.2n by utilizing 5.875 Ma for
the top of C3A.1n, the radiometric age of Baksi [1992] for
the top of C5n.1n (9.64 Ma), and the seafloor distances given by
Cande and Kent [1992a].
De Boer [1982] and Schwarzacher and Fischer [1982]
proposed that bundling of carbonate couplets (in a nearly 5:1
ratio) in Barremian-Cenomanian (middle Cretaceous) Italian marly
limestones represents precessional forcing grouped into 100 kyr
increments by the eccentricity envelope. Herbert et al.
[1994] combined cyclostratigraphic estimates of the duration of the
Aptian, Albian and Cenomanian with the Obradovich [1993] age
for the Cenomanian-Turonian boundary (93.5 
0.2 Ma) to estimate
ages for the Barremian-Aptian, Aptian-Albian and Albian-Cenomanian
stage boundaries (Table 2).