Page 2 - 23

ARTICLE IN PRESS
1568 K. Lambeck et al. / Quaternary Science Reviews 23 (2004) 1567–1598
established (Lambeck and Johnston, 1995; Lambeck
and Bard, 2000). Apart from tectonic signals the
dominant contribution to sea-level change in the
Mediterranean is the eustatic part. However, while
the region lies well beyond the margins of the former
large ice sheets glacio-isostatic effects are significant.
The important consequence of the large ice sheets is that
for the entire length of Italy the crust is subsiding in
response to the last deglaciation. (We ignore here any
contributions from the Alpine glaciation. Simple order
of magnitude estimates indicate that their magnitudes
are smaller than those arising from the Eurasian and
North American ice sheets, but that they may contribute
in a minor way to subsidence of the crust along the
North Adriatic coast. However see note added in proof
below) The hydro-isostatic contribution results from the
loading of the ocean floor by the meltwater entering into
the Mediterranean and causes maximum subsidence
within the central parts of the basin and relative uplift at and subsidence are representative of shorter time
the coast and inland. If the earth-rheology, the history
of the global ice sheets, and the geometry of the ocean
basins are known, then both isostatic effects can be
evaluated. If the relevant parameters are only partly
known, then observations of sea-level change can be
used to improve the model parameters. Thus on the one
hand the model can be effectively used as a formulation
for interpolating between fragmentary observations of
sea-level change across the region. But on the other
hand, we also wish to have a model that yields
parameters that are physically insightful about the earth
rheology and the ice sheets.
Tectonically the Italian peninsula cannot be consid-
ered to be stable on Quaternary and longer timescales. It
comprises B8000 km of coastline of which approxi-
mately 25% consists of low-lying plains, some of which
are subsiding with a potential for flooding (Antonioli
and Leoni, 2000; Antonioli et al., 2001a). Other areas, in
particular the southern coastal zone of Calabria and
eastern Sicily, are subject to uplift at rates reaching
1–1.4 mm yearÀ1 (Monaco et al., 2001; Antonioli et al.,
2003; Miyauchi et al., 1994). There are marked
differences in local tectonic evolution between adjacent
coastal zones: areas exhibiting an unstable behaviour
even in recent times occurring adjacent to, and alter-
nating with, areas that have been virtually stable since
the middle Pleistocene. Other areas, such as the coasts of
Tuscany, Sardinia, Campania and parts of Latium
appear to have been stable for much of the later part
of the Quaternary. Thus any analysis of the Italian sea-
level data in terms of eustasy and isostasy must consider
the possibility of vertical tectonic movement along the
entire coastal sector. In this regard, an important
observation in Italy and elsewhere in the Mediterranean
is the elevation of the Last Interglacial (marine isotope
stage MIS 5.5 or 5e) sea level. Here this level (named
also Tyrrhenian) is characterized by several morpho-
and litho-stratigraphic markers including prominent
notches, lagoonal sedimentary facies, fossil beaches
and terraces that often contain a typical warm-fossil
association named Senegalese fauna including Strombus
bubonus (Gignoux, 1913). This gastropod is found in the
central and eastern Mediterranean Sea only at the time
of the Last Interglacial highstand and it provides a
unique marker of the MIS 5.5 horizon. Globally, in
tectonically stable regions, Last Interglacial shorelines
lie a few meters above present sea level (e.g. Stirling
et al., 1998) but in some locations in Italy, such as
Calabria, it occurs at more than 100 m. Elsewhere, such
as along the northern shores of the Adriatic Sea, it
occurs below present sea level. We use the position of
this marker as an indicator of tectonic stability or
otherwise and in the latter case we make a first-order
correction for the tectonic contribution based on the
assumption that long-term (105 years) rates of uplift
intervals (B103 years). If adequately representative
models can be developed then this assumption can be
tested a posteriorily.
In this paper we first summarize the types of evi-
dence available for sea-level change along the coast of
Italy (Section 2) and discuss the data and tectonic
stability for each location (Section 3). We next
summarize the eustatic–isostatic model and its uncer-
tainties (Section 4) and compare the model predictions
with the observational evidence (Section 5) for stable,
uplifting and subsiding sites. In Section 6 some of the
nominal model parameters are re-evaluated based on the
Italian data, primarily the global eustatic sea-level
function and some local estimates of vertical crustal
motion where the underlying assumption of uniformity
of uplift throughout the last interglacial cycle does not
appear to be valid. With these revisions, predictive
models for the relative sea-level change and shoreline
evolution across the region are presented for the interval
from the Last Glacial Maximum (LGM) to the present.
Section 7 summarizes the main results.
2. Sea-level indicators and their accuracies
A range of different sea-level indicators have been
used which can be divided into biological, sedimentolo-
gical, erosional and archaeological categories. The last
have been discussed for the Mediterranean as a whole by
Flemming (1969) and for some Italian locations by
Schmiedt (1972) and Pirazzoli (1976) and we do not
discuss the methodologies in any detail. Some of the
biological evidence from the Italian region have received
less attention and we comment briefly on these sea-level
markers here. Most sea-level indicators refer to some
specific part of the tidal range and while this is generally
small in the Mediterranean, corrections to mean sea