Page 3 - Vacchi_Marriner_2016

174 M. Vacchi et al. / Earth-Science Reviews 155 (2016) 172–197

Fig. 1. Spatial extent of the study area. Numbered rectangles denote the location of sea-level data for this paper, grouped into regions as explained in the text. GS is Gibraltar Strait, CO is
Corsica, SA is Sardinia, SI is Sicily.

2. Study area 2.1. Original source of the data
The study area encompasses the western Mediterranean coast at lat- The post-LGM RSL changes in the western Mediterranean have been
itudes comprised between 45.7°N and 33°N and longitudes comprised investigated since the late 1960s (e.g., Flemming, 1969; Pirazzoli, 1976;
between −0.3°E and 16.2°E (Fig. 1). The data were collected in central Lambeck and Bard, 2000; Lambeck et al., 2004a; Pirazzoli, 2005;
and northern Spain, southern France and Corsica, much of the Italian Antonioli et al., 2009; Anzidei et al., 2014). In particular, several studies
coast (including Sicily and Sardinia), Malta, Slovenia, northern and cen- have focused on the late Holocene (last 4 ka) using geo-archaeological ar-
tral Croatia and southern Tunisia (Fig. 1). chives to reconstruct past sea-level histories in areas of Spain
Western Mediterranean tides have an average amplitude of about (e.g., Carmona-González and Ballester, 2011), France (e.g., Morhange
0.4 m, although they do show spatial variability based on coastal geom- et al., 2001, 2013), Sardinia (e.g., Antonioli et al., 2007; Orrù et al., 2011,
etry and bathymetry (Tsimplis et al., 1995; Antonioli et al., 2015). Tidal 2014), Croatia (e.g., Antonioli et al., 2007; Faivre et al., 2010), Tunisia
amplitude is very small near the amphidromic points but it may reach (e.g., Anzidei et al., 2011), Sicily (e.g., Scicchitano et al., 2008) and Malta
amplitudes of up to 2 m in the Gulf of Gabes and the North Adriatic (e.g., Marriner et al., 2012b; Furlani et al., 2013). One of the main goals
Sea (Tsimplis et al., 1995; Marcos et al., 2009). Around the Straits of Gi- of the geoarcheological studies has been to establish the RSL histories dur-
braltar, the Atlantic Ocean affects Mediterranean tides in the Alboran ing the first millennium BC (e.g., Pirazzoli, 1976; Morhange et al., 2001;
Sea (Fig. 1), but its influence rapidly declines further east (e.g., Marcos Lambeck et al., 2004b; Goiran et al., 2009; Evelpidou et al., 2012). In
et al., 2009). Three main deltas, the Ebro, the Rhone, and the Po, repre- Marseille's ancient harbor, Morhange et al. (2001), coupled geo-
sent the main sedimentary inputs into the western Mediterranean basin archeological and biological sea-level data to reconstruct RSL changes
(Maselli and Trincardi, 2013; Anthony et al., 2014). since the Bronze Age. They proposed that RSL was at ~−1.5 m in the
From a geological point of view, the western Mediterranean is a tec- Middle Bronze Age (~3.6 ka BP) and ~−1.2 m in the Late Bronze Age
tonically complex area where two small oceanic basins (the Tyrrhenian (~2.8 ka BP). Between the Archaic and Roman periods (~2.0 to
and Liguro-Provençal back-arc basins) occur along the Nubia–Eurasia 1.5 ka BP), RSL ranged between −0.8 and −0.4 m with respect to the cur-
convergent margin and are separated by the Corsica–Sardinia rigid con- rent Mean Sea Level (MSL).
tinental block (Fig. 2, Jolivet and Faccenna, 2000; Jolivet et al., 2008; An ongoing debate concerns the archeological interpretation of
Faccenna et al., 2014). ancient Roman fish-tanks (piscinae), with implications for the recon-
In the western Mediterranean, some parts of the back-arc basin mar- struction of RSL during the Roman period. These coastal structures
gins are undergoing compressional tectonics (e.g., Billi et al., 2011). This are assumed to be among the most reliable archeological RSL indica-
tectonic regime is particularly active, from west to east, along the east- tors (Lambeck et al., 2004b; Auriemma and Solinas, 2009; Morhange
Alboran, Algerian, and south-Tyrrhenian margins (Billi et al., 2011; and Marriner, 2015) because sea level and fish-tank architecture are
Faccenna et al., 2014; Fig. 2). In the central Mediterranean, the contrac- closely correlated and their chronology is well constrained to be-
tional orogen swings around the Adriatic Sea and marks the deformed tween ~1st century BC (~2.1 ka BP) and ~1st century AD
margins of Adria (Fig. 2). (~1.9 ka BP).