Page 9 - slootman_et_al_2016
P. 9

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

Table 2
Time comprised within the depositional system, backset-bedded clinoform units and dune cross-bedded clinoform units, and the average recurrence period of backset-
generating events and dune-activating events.

Depositional system                                           Unit     Lower limit  68% CI2  Mean  Upper limit  95% CI1
Time comprised within studied clinoform succession            kyr      95% CI1      364      458   68% CI2      705

     (20 clinoform couplets)                                           287                         573

Backset-bedded clinoform units                                kyr 14.3              18.2     22.9  28.6         35.3
Average time comprised within single clinoform couplet, i.e.

     average recurrence period backset-generating events

Dune cross-bedded clinoform units                             yr 287 364 458 573 705
Average recurrence period, n = 50 dune-activating events      yr 29 36 46 57 71
Average recurrence period, n = 500 dune-activating events

1,2 CI: Confidence Interval. Lower and upper limits of 95% (68%) confidence interval correspond to 47.5% (34%) on either side of the mean at 2.5% (16%) and 97.5% (84%)
cumulative probability density. See Fig. 3B for visual representation.

where the critical Shield stress required for the initiation of mo-    to ca. 30 m thick (e.g. Massari and D’Alessandro, 2012), are par-
                                                                       ticularly preserved between 2.5–2.1 Ma and between 1.6–0.8 Ma
tion is τcr∗ = 0.047 for well sorted, fine gravel and θ is the          (Catalano et al., 1998).

non-dimensional mobility parameter (Van Rijn, 1984). The latter            Although the clinoform succession of present-day Favignana is
                                                                       incomplete and extended NW towards the carbonate factory, one
ranges from 0.15 to 1.5 for subaqueous dunes in the observed           can argue that, within error, the onset of major progradation of the
                                                                       Favignana carbonate ramp correlates to the important sequence
grain size distribution (Van den Berg and Van Gelder, 1993). How-      boundary at 1.6 Ma, which formed as the result of a eustatic
ever, values above θ = 0.5 yield unrealistically high flow velocities   lowstand (Wornardt and Vail, 1991; Catalano et al., 1998). The
(>3.0 m/s) when compared to modern environments (e.g. Kubicki,         resulting global regression created wide-spread hiatuses in the on-
2008; Barrie et al., 2009) and were therefore not considered in the    shore basins of, in particular, western Sicily (Catalano et al., 1998),
                                                                       which apparently coincided with the commencement of carbon-
calculations. The non-dimensional sediment flux qs∗ is dimension-       ate production in the Favignana system. On the basis of available
alised by                                                              data, it is tentatively suggested that the major sea-level lowstand
                                                                       at ca. 1.6 Ma placed the submarine high between Favignana and
qs = qs∗ D50   ρsed − ρamb  g D50                             (5)      Levanzo in a position favourable to the onset of heterozoan carbon-
                  ρamb                                                 ate production. In the subsequent 0.3–0.7 Myr, the combination of
                                                                       aggradation, tectonic uplift and long-term sea-level rise apparently
where ρsed and ρamb are the densities of the sediment                  maintained such conditions whilst preventing subaerial exposure
                                                                       of the carbonate factory during high-frequency lowstands. This re-
(2700 kg/m3) and the ambient water. Cumulative flow duration            sulted in the continuous progradation of the Favignana carbonate
for dune cross-bedded clinoform units is estimated by dividing         ramp in the absence of any recognisable obliquity-forced signature.
the dip-parallel, cross-sectional area of the clinoform unit A by
sediment flux qs. It follows that for the deposition of one dune        5.2. Decennial to centennial storms
cross-bedded clinoform unit of dimensions similar to those of
the examined backset-bedded clinoform unit, the cumulative flow             The formation and migration of subaqueous dunes in present-
duration is 1.25 ± 1 yr (Table 1). This implies that individual dune-  day shelf environments result from strong bottom currents (0.8–
activating events, of which 50 to 500 together generated a single      1.9 m/s, e.g. Kubicki, 2008; Barrie et al., 2009) related to tides
dune cross-bedded clinoform unit, lasted between approximately         (Barrie et al., 2009), strong winds (Kubicki, 2008; Barrie et al.,
5 hours and 2 weeks (Table 1).                                         2009), sinking dense water masses (Lykousis, 2001) or ocean wa-
                                                                       ter circulation patterns (Flemming, 1978). Numerous studies deal
5. Discussion                                                          with modern subaqueous dunes that are active exclusively dur-
                                                                       ing extreme storms (e.g. Bassetti et al., 2006; Li and King, 2007;
5.1. Palaeogeography of the Favignana carbonate ramp                   Kubicki, 2008; Lo Iacono et al., 2010). Numerical modelling by
                                                                       Bassetti et al. (2006) on the outer shelf of the Gulf of Lions (west-
    The age constraints produced here enable the reconstruction of     ern Mediterranean) shows that a single day of 20 m/s wind, with
the system dynamics of the Favignana carbonate ramp, thereby re-       a recurrence period of 4 yr, is sufficient for the erosion and trans-
solving its place in the framework of the Pleistocene sequences        port of fine-grained sand at 100 m water depth. According to these
of Sicily (Catalano et al., 1998). The major source of skeletal        authors, there is “no doubt that much more severe meteorological
hash was almost certainly a vast carbonate factory on the in-          conditions allow episodic transport of all classes of sand and reju-
ferred sill between the tectonic ridges of the islands of Favignana    venation of subaqueous dunes”.
and Levanzo, NW of the studied clinoform succession as hinted
by the SE ramp slope progradation and mean dip of subaque-                 The calculated flow velocities for the formation of subaqueous
ous dune cross-beds. The statistically robust age model, developed     dunes in the Favignana system (1.4–2.7 m/s) are comparable to the
from strontium isotope stratigraphy, brackets the period of ma-        range of values measured in modern environments. The Aegadian
jor clinoform progradation between 1.6 ± 0.1 and 1.1 ± 0.1 Ma          Archipelago is at present not affected by significant tidal currents,
(95% confidence level). The time comprised within the Favignana         nor by the movement of large water masses, similar to the con-
system, spanning several hundred kyr, refutes a 41-kyr orbital         ditions that are believed to have prevailed throughout the Early
obliquity-controlled genesis as proposed for the clinoform suc-        Pleistocene. The influence of strong winds in the modern Mediter-
cessions of the Lower Pleistocene of Sicily (Catalano et al., 1998;    ranean region, on the other hand, was assessed by Trigo et al.
Massari and D’Alessandro, 2012). This is in line with the gener-
ally greater length, height and thickness of the Favignana clinoform
units compared to those of the carbonate ramp successions of
Sicily. The latter, forming well-developed carbonate sequences up
   4   5   6   7   8   9   10   11   12   13