Geology Reference
In-Depth Information
et al
., 1974; Chappell & Shackleton, 1986). The
frequencies of these oscillations match those of
the precession and eccentricity cycles, and they
provide a driver for the bundling of depositional
cycles that is fundamentally linked to carbonate
sedimentation.
Additional sedimentological evidence for
Milankovitch band forcing of depositional cycles
relates to the rates at which complete carbonate
cycles form, particularly vadose diagenetic facies.
While the cycles measured at Mendola Pass do
not have diagenetic caps, many of the cycles in
the equivalent Latemar succession do have well-
developed diagenetic caps. The presence of these
diagenetic caps must be accounted for in a tem-
poral sense for the Latemar cycles and in a time-
correlative sense for Mendola cycles. If Holocene
examples are used as a temporal benchmark, rapid,
millennial-scale sea-level oscillations do not pro-
vide adequate time for observed vadose diagenetic
features to form. Radiocarbon-calibrated rates of
Holocene caliche formation average 2-4 cm kyr
1
(James, 1972; Robbin & Stipp, 1974; Handford
et al
., 1984; Demicco & Hardie, 1994), and several
studies suggest even lower rates, at 0.4-1 cm kyr
1
(Lucia, 1968; Davies, 1970; Evamy, 1973). The
caliche that caps Latemar cycles ranges in thick-
ness from 1 to 30 cm, averaging 6 cm in thickness
(Goldhammer
et al
., 1987). In order for diagenetic
caliche caps to form within a millennial cyclic
framework (entire cycle in 0.9-1.97 kyr), rates of
caliche formation must operate at least an order
of magnitude faster (20-40 cm kyr
1
), if not more,
than those in the Holocene. Additionally, stud-
ies of Holocene tepees at Lake MacLeod, Western
Australia by Handford
et al
. (1984) show that
sediment buckles and is tilted into tepee antiforms
at an average rate of 4 cm kyr
1
, and void-fi lling
aragonite cements within tepees grow at extremely
slow rates of 0.2-0.4 cm kyr
1
. Goldhammer
et al
.
(1987) and Hardie
et al
. (1991) reported tepee
zones in the Latemar successions with thicknesses
varying from 1 to 13 m, and individual cement
bands up to 20 cm thick. Individual tepee anti-
forms are typically onlapped by sediments from
subsequent cycles, suggesting that tepees were
completely formed prior to the deposition of the
next cycle(s). If Holocene rates of tepee formation
from Lake Macleod are even close approximations
of rates of tepee development in the Latemar, the
multimetre tepee structures at Latemar section
alone represent several hundred thousand years
of time. In order for these same tepee structures to
have formed caps on millennial cycles, both tepee
antiforms and cements would have to form at rates
that are at least two orders of magnitude faster
(40-400 cm kyr
1
antiform; 20-40 cm kyr
1
cement)
than those documented at Lake MacLeod.
In addition to sedimentological evidence,
spectral analysis of Latemar cycle thickness
and rank series indicates a close match between
bundling frequencies of Latemar cycles and
Milankovitch orbital cycles. The clarity of the
record and the non-trivial nature of the match
is especially apparent in studies using rank series
analysis, where plots of depth-ranked units
appear to have formed in lock-step synchroneity
with Milankovitch cycle insolation changes (see
Fig. 8 of Preto
et al
., 2004). However, tuning the
Latemar cycles to a sub-Milankovitch component
(i.e. 4.2 kyr; Zühlke
et al
., 2003) also preserves
Milankovitch frequency components, but at a dif-
ferent order (see Figs 11 and 12 and Table 2 of
Zühlke, 2004), pointing to a 'fractal-like' aspect of
the different orders of the Milankovitch rhythms
(fi rst noted by Schwarzacher, 1998). The differ-
ence lies in the assignment of the precession:
the pure Milankovitch interpretation suggests
that the precession has driven fundamental cycle
formation (i.e. 20 kyr Latemar cycles and 100 kyr
megacycles), while the mixed Milankovitch
scale and sub-Milankovitch scale interpretation
suggests that precession drove the formation of
megacycles (i.e. 4 kyr Latemar cycles and 20 kyr
megacycles).
In light of this 'fractal' equivalence, the mixed
Milankovitch and sub-Milankovitch model
deserves closer attention. Thus far, no specifi c
proposal has been offered for the required eustatic
driver operating at a millennial scale of around
4 kyr. However, it is notable that recent work on
Pleistocene eustasy may lead to the identifi ca-
tion of such a driver. Chappell (2002) reported
6 kyr recurrent coral terraces uplifted in the Late
Pleistocene of Papua New Guinea suggestive of
sea-level oscillations with amplitudes as large as
15 m. Possibly this is the analogue to the 4.2 kyr
Latemar driver. There is also evidence of sim-
ilar eustatic behaviour in the uplifted Pleistocene
coral terraces of Barbados (Thompson & Goldstein,
2005). However, both locations are situated in tec-
tonically active areas, involving rapid uplift of
the islands related to the geodynamics of adjacent
convergence zones. In particular, the rather large
(>10 m) amplitudes that have been suggested for
these millennial eustatic changes depend on the
accuracy of the uplift rates that have been estim-
ated for these islands. It is also important to
