258
w. Schwarzacher
If a sedimentary sequence is predominantly determined by water depth, then
the large fluctuations in sea level will clearly produce much better defined sedimentary cycles than the small fluctuations associated with Milankovitch cycles.
One can argue, in a similar way, that any change of paleodepth will be less noticeable in deep water than in shallow water, although this is only partly true.
Processes like sediment shedding, control of nutrients or transparency may
originate in coastal areas and can have a considerable influence on deep water
sedimentation. Nevertheless, the best environments for studying the interaction
between sequences and Milankovitch cycles would appear to be shallow areas in
which sequence boundaries are still developed but where sedimentation is continuous enough to record the Milankovitch cycles.
Of particular interest are studies of peritidal carbonate shelves and platforms.
In these environments, sedimentation is almost exclusively determined by the
available accommodation space, which is determined by the paleo water depth.
If subsidence is constant, paleo water depth can be estimated by subtracting a
constant trend from the cumulative sediment thickness curve (Schwarzacher
1975). Such Fischer plots have been used to reconstruct 3rd-order sea-level
changes in cyclic Ordovician carbonates (Read and Goldhammer 1988) and similarly in Devonian carbonates (Elrick 1995). The time represented by the individual cycles in both studies is not definitely known, but they are believed to be
of Milankovitch origin and therefore can be used for relative timing. The 3rd-order sea-level changes are found to be of the order of 20 m in amplitude but they
are quite irregular and the intervals between successive high stands vary between 17 and 40 cycles. Errors can be introduced in this timing by incomplete
sedimentation (missed beats) but the evidence strongly suggests that the 3rd-order fluctuations are nowhere near time-periodic.
A definite attempt to time sequences with Milankovitch cycles has been made
by Strasser (1991). Cycles in a lagoonal to supratidal carbonate sequence oflate
Tithonian to early Berriasian age in the French Jura Mountains have been tentatively identified as 20-ka, 100-ka and 400-ka Milankovitch cycles. Furthermore,
two sequence boundaries from the Haq et al (1987) chart could be recognised.
The interval which, according to the global chart is 2.8 Ma, works out to be 3.6
Ma, according to the cyclostratigraphic timing. However, Strasser stresses the
uncertainty not only in identifying the sequence boundaries, but also the uncertainty of having identified the Milankovitch cycles correctly. Additional sequence boundaries are suggested by Strasser and these are placed at intervals of
four or at multiples of four 100-ka cycles. The status of these subsidiary sequences is not clear from his paper, as the 100-ka cycles are also referred to as sequences. Similar effects of sea-level fluctuations on the development of Milankovitch
cycles have also been observed in the Swiss Jura Mountains (Pittet 1994).
An important new method of analysing stratigraphic sections is based on
wavelet transforms which can provide frequency spectra of very short time intervals (Bolton et a1.1995; Prokoph and Barthelmes 1996).The method is ideally
suited to locating changes in cyclicity which often coincide with sequence
boundaries (B.Niebuhr 1996, pers.com).
w. Schwarzacher
If a sedimentary sequence is predominantly determined by water depth, then
the large fluctuations in sea level will clearly produce much better defined sedimentary cycles than the small fluctuations associated with Milankovitch cycles.
One can argue, in a similar way, that any change of paleodepth will be less noticeable in deep water than in shallow water, although this is only partly true.
Processes like sediment shedding, control of nutrients or transparency may
originate in coastal areas and can have a considerable influence on deep water
sedimentation. Nevertheless, the best environments for studying the interaction
between sequences and Milankovitch cycles would appear to be shallow areas in
which sequence boundaries are still developed but where sedimentation is continuous enough to record the Milankovitch cycles.
Of particular interest are studies of peritidal carbonate shelves and platforms.
In these environments, sedimentation is almost exclusively determined by the
available accommodation space, which is determined by the paleo water depth.
If subsidence is constant, paleo water depth can be estimated by subtracting a
constant trend from the cumulative sediment thickness curve (Schwarzacher
1975). Such Fischer plots have been used to reconstruct 3rd-order sea-level
changes in cyclic Ordovician carbonates (Read and Goldhammer 1988) and similarly in Devonian carbonates (Elrick 1995). The time represented by the individual cycles in both studies is not definitely known, but they are believed to be
of Milankovitch origin and therefore can be used for relative timing. The 3rd-order sea-level changes are found to be of the order of 20 m in amplitude but they
are quite irregular and the intervals between successive high stands vary between 17 and 40 cycles. Errors can be introduced in this timing by incomplete
sedimentation (missed beats) but the evidence strongly suggests that the 3rd-order fluctuations are nowhere near time-periodic.
A definite attempt to time sequences with Milankovitch cycles has been made
by Strasser (1991). Cycles in a lagoonal to supratidal carbonate sequence oflate
Tithonian to early Berriasian age in the French Jura Mountains have been tentatively identified as 20-ka, 100-ka and 400-ka Milankovitch cycles. Furthermore,
two sequence boundaries from the Haq et al (1987) chart could be recognised.
The interval which, according to the global chart is 2.8 Ma, works out to be 3.6
Ma, according to the cyclostratigraphic timing. However, Strasser stresses the
uncertainty not only in identifying the sequence boundaries, but also the uncertainty of having identified the Milankovitch cycles correctly. Additional sequence boundaries are suggested by Strasser and these are placed at intervals of
four or at multiples of four 100-ka cycles. The status of these subsidiary sequences is not clear from his paper, as the 100-ka cycles are also referred to as sequences. Similar effects of sea-level fluctuations on the development of Milankovitch
cycles have also been observed in the Swiss Jura Mountains (Pittet 1994).
An important new method of analysing stratigraphic sections is based on
wavelet transforms which can provide frequency spectra of very short time intervals (Bolton et a1.1995; Prokoph and Barthelmes 1996).The method is ideally
suited to locating changes in cyclicity which often coincide with sequence
boundaries (B.Niebuhr 1996, pers.com).