Page 5 - slootman_et_al_2016
P. 5

196 A. Slootman et al. / Earth and Planetary Science Letters 444 (2016) 192–204

Fig. 3. Results of strontium isotope stratigraphy of an interval comprising 20 clino-  The measured time span for deposition of the studied succession
form couplets within the Pleistocene carbonate ramp succession of Favignana Island.    is controlled by the confidence level; larger confidence intervals
(A) 87Sr/86Sr values of thirteen fossil rhodolith samples (Table S1) plotted on the    result in upper and lower limits farther removed from the me-
LOWESS best-fit curve (Skeleton Version 5) for strontium isotope ratios as a func-      dian duration of 458 kyr. On a 95% confidence level, the duration
tion of time (thick line: mean age; thin lines: limits of the 95% confidence interval)  of deposition of all samples is between 287 and 705 kyr, a period
calibrated by McArthur et al. (2012). Error bars show the 95% confidence interval       that began between 1.71 and 1.51 Ma and ended between 1.22 and
for each sample corresponding to two standard deviations. (B) Cumulative proba-        1.00 Ma (Fig. 3B), confirming the assumed Lower Pleistocene age
bility of lower and upper age limits of the sampled stratigraphic interval. A larger   (S´ la˛czka et al., 2011). This time span is larger than that of the typ-
confidence interval corresponds to wider ranges of the lower and upper age limit,       ical clinoform successions of the Lower Pleistocene of Sicily (ca.
thereby increasing the possible range of the time contained within the sampled cli-    41 kyr, Catalano et al., 1998; Massari and D’Alessandro, 2012), im-
noform succession (i.e. duration). For example, on a 95% confidence level, deposition   plying a deviating evolution of the Favignana system compared to
of the sampled stratigraphic interval lasted between 287 and 705 kyr, commencing       the regional standard model. Since the sampled succession con-
between 1.71 and 1.51 Ma and ending between 1.22 and 1.00 Ma.                          tains 20 clinoform couplets, deposition of each clinoform couplet
                                                                                       covered an average time span of 14 to 35 kyr, which thus repre-
mobility parameters (sensu Van Rijn, 1984) up to ca. 25% lower                         sents the average recurrence period of backset-generating events
than those typical for siliciclastic particles. The mobility difference                and the average time comprised within a dune cross-bedded cli-
between carbonate and siliciclastic sands is, however, less than the                   noform unit.
precision in the lower ranges of the available siliciclastic models.
Hence, use of the above methods is justified, as the outcome of                         4.2. Anomalous backset-bedded clinoform units
the present study is not significantly affected by the difference in
hydrodynamic behaviour between siliciclastic and carbonate parti-                      4.2.1. Sedimentary features
cles.                                                                                      The backset-bedded clinoform units outcropping in the large

4. Results                                                                             sea cliff (Fig. 2) display very similar architecture and facies, despite
                                                                                       their variation in thickness. They consist invariably of amalgamated
4.1. Strontium isotope stratigraphy                                                    sets of backset-bedding. The lower boundary of backset-bedded
                                                                                       clinoform units is typically increasingly erosive towards the prox-
    The thirteen fossil rhodolith samples from a succession of 20                      imal ramp slope, where these units are up to 10 m thick (Fig. 2).
clinoform couplets range in age from 1.500 to 1.230 Ma (Fig. 3A,                       Large variations in grain size occur throughout the units, ranging
Table S1 in Supplementary Material). Due to an average uncer-                          from sand to gravel. The near absence of bioturbation empha-
tainty of ca. ±170 kyr (95% confidence level) on the age of in-                         sises the event character and rapid deposition of backset-bedded
dividual samples, mean ages do not plot in the chronological order                     clinoform units, with escape burrows (Ophiomorpha) being rarely
inferred from their stratigraphic position. However, the objective of                  present in the uppermost interval. Although individual backset-
87Sr/86Sr dating was not to constrain individual sample ages, but                      bedded sets formed by the scouring and filling associated with
to investigate the time contained within the clinoform succession.                     a single hydraulic jump (cf. Massari, 1996), the generation of a
                                                                                       backset-bedded clinoform unit is related to the repetitive forma-
                                                                                       tion of numerous hydraulic jumps during a single sediment gravity
                                                                                       flow event, in which part of the backset-dominated stratification
                                                                                       passes from one set into the other generating an amalgamated
                                                                                       stacking pattern of backset-bedded sets (cf. Cartigny et al., 2014).
                                                                                       The backset-bedded clinoform units together comprise about half
                                                                                       of the thickness of the studied succession (Fig. 2).

                                                                                           One backset-bedded clinoform unit with high exposure qual-
                                                                                       ity and easy accessibility was investigated in more detail. This
                                                                                       unit dips slightly oblique to the outcrop and parallel to clinoform
                                                                                       progradation and palaeotransport. It is up to 7 m thick in the stud-
                                                                                       ied section and can be traced over 400 m updip (Figs. 2 and 4;
                                                                                       high-resolution photo available through Supplementary Material).
                                                                                       The lower part consists of a set of coarse sand- to granule-grained
                                                                                       backset-beds, which dip gently upstream and become progressively
                                                                                       wavier upwards. The contact with the overlying backset-bedded
                                                                                       set with gravel-sized bioclasts is erosive, culminating in a 3 m-
                                                                                       deep, asymmetric scour filled with backset-bedding, in which each
                                                                                       subsequent backset-bed truncates the preceding one progressively
                                                                                       farther upstream, generating a composite erosion surface and an
                                                                                       apparent onlap onto the upstream side of the scour (Fig. 5).

                                                                                       4.2.2. Palaeohydraulic reconstruction
                                                                                           The preserved wavy geometries are explained as the result of

                                                                                       progressively steepening antidunes that formed on the stoss-side
                                                                                       of a larger bedform (Figs. 5B–C), generated by a gravity flow rush-
                                                                                       ing down the ramp slope. The combination of severe erosion and
                                                                                       deposition of backset-stratification, points to hydraulic jumps and
                                                                                       breaking waves (Figs. 5D–G), as expected in steepening antidune
                                                                                       sequences (cf. Kennedy, 1963; Hand, 1974; Cartigny et al., 2014).
                                                                                       Mean flow velocity U is related to antidune wavelength L (Hand,
                                                                                       1974) by
   1   2   3   4   5   6   7   8   9   10