Page 4 - 23

ARTICLE IN PRESS
1570 K. Lambeck et al. / Quaternary Science Reviews 23 (2004) 1567–1598
considerably increased, reaching as much as 10 m, unless
the samples are from near the lower part of the marsh or
peat deposit or unless compaction corrections are
possible. This information is not always available in
the older publications of sea-level data.
2.1.6. Fossil sand beach
In some localities, particularly the uplifting areas of
southern Italy, fossiliferous sands have been identified.
These contain gastropods and bivalves (Bolma rugosa,
Osilinus turbinatus, Hesaplex trunculus species) that
form biological associations characteristic of a lower
mesolittoral cliff environment subjected to regular
periods of immersion. They may also contain species
(Ostrea, Spondilus, Cerastoderma Cardium) that do not
show relevant ecological associations. Similar mixed
associations occur along the modern infrashoreline zone
at the same localities as the fossil outcrops. The error
bars of these results could be as large as 10 m. Where
possible, the same methodology as described for
the lithophaga deposits has been used to reduce the
uncertainty of the depth range of these deposits (i.e.
using scuba-dive transect information) (Antonioli et al.,
2003).
2.1.7. Beach rock
Beach rock is constituted from clastic shoreline
deposits cemented by calcitic-magnesitic or aragonitic–
carbonates in or near the intertidal zone, often at the
interface of the freshwater–marine phreatic flow. This
precipitation can be exceptionally fast, of the order of 10
years and, depending on local conditions, this interface
may occur above or below mean sea level. In particular,
supratidal cementation by carbonate cements is com-
mon and much debate still exists about the uppermost
level of cementation. Also, dating the time of formation
is associated with difficulties, and ages may be either too
old because of contamination from older carbonate or
they may be too young because of post-depositional
alteration of the cement. Hopley (1986), in reviewing the
suitability of beachrock as a sea-level marker, concludes
that it is an unreliable recorder of sea level and while his driven by sediment loading.) In southern Italy the MIS
discussion deals with low-latitude deposits we adopt a
cautious approach in its use here. Numerous beach rock
deposits have been preserved at different depths along
the Sardinia-Corsica continental shelf with remarkable
continuity of the outcrops (De Muro and Orru" , 1998).
The thickness of the sampled deposits is mostly between
4 and 5 m and we provisionally adopt the error bars of
+1 and À5 m for these results as given by De Muro and
Orru" . This encompasses the likely uncertainty of the
relation between formation and sea level.
2.1.8. Submerged archaeological remains
Flemming (1969), Schmiedt (1972), Caputo and Pieri
(1976), and Pirazzoli (1976) have discussed the archae-
ological evidence for sea-level change in considerable
detail. Different archaeological markers provide sea-
level estimates with different accuracies and it is not
possible to assign the same error bars to all data. For
estimating the altitude of the Roman piscinae, for
example, the architectural feature that relates directly to
sea level must be identified and is not necessarily the
same for all holding tanks (cf. the piscinae of Torre
Astura and the fish tanks from Briatico). Such observa-
tions are corrected for tidal effects with an uncertainty
of 70.20 m.
3. Observed sea-level change and tectonic uplift
Table 1 summarizes the evidence for change in the
land–sea intersection at the 30 Italian sites. Data from
Tunisia has also been included because this site increases
the north–south extent of the region and potentially
provides information for testing the isostatic models for
glacio-hydro-isostatic rebound. Site locations are given
in Fig. 1. The MIS 5.5 (or 5e) shoreline is well developed
along long sections of the Italian coast and its present
elevation is used to establish whether significant tectonic
uplift or subsidence has occurred because, in tectonically
stable regions and away from former ice sheets, its
position is typically a few meters above present sea level
(Stirling et al., 1998). Along the northern and central
Tyrrhenian coast (southern Tuscany, northern and
southern Latium, Campania, Sardinia) the marine
notches and inner margin terraces corresponding to
the MIS 5.5 highstand occur from within a few meters
up to 4–10 m above present sea level (Hearty and Dai
Pra, 1986; Antonioli et al., 1999b; Carobene and Pasini,
1982). (An uncorrected uniform uplift of 10 m since the
Last Interglacial introduces an error of B1 m in the
elevation of early Holocene sea levels.) Thus here
the tectonic movement since the Last Interglacial
appears to have been small and the Holocene data from
these localities is assumed to be largely free from vertical
tectonic movement. (We distinguish between movements
of the crust driven by various tectonic processes other
than the isostatic response to the ice–ocean loading. The
tectonic component includes, for example, subsidence
5.5 highstand reaches some of the highest altitudes in the
Mediterranean: 175 m in Sicily (Monaco et al., 2001)
and 162 m in Calabria (Miyauchi et al., 1994). In
contrast, along the Adriatic coast, particularly the
central and northern parts, considerable tectonic sub-
sidence has occurred because coastline features attribu-
table to the MIS 5.5 stage are found only in cores at
depths of À100 m to À120 m below present sea level
(Amorosi et al., 1999). Fig. 1 summarizes the informa-
tion on the elevation of the MIS 5.5 shoreline in Italy.
The tectonic vertical movement and its uncertainty