Tyrrhenian sea level at 2000 BP: evidence from Roman age fish tanks and their geological calibration

Abstract

Following the pioneering work of Schmiedt et al. (1972) on establishing the level of the Tyrrhenian Sea in Antiquity, a number of studies have examined this evidence from Roman Period fish tanks but with significantly different outcomes due primarily to different interpretations of the functional level of these pools at the time of their construction. As part of a longer term project to understand the causes of sea-level change around the Italian coast, we have reexamined and resurveyed 12 well-documented fish tanks, all based on the same construction principles, from the Tyrrhenian coast (between Formia and Orbetello) for which it can be established that they were in open contact with the sea at the time of operation. The structural features that tidally control the exchange of water used to define the ancient local sea level are identified as the channel thresholds, the sluice gate and sliding post positions, and the lowest level crepido. These are consistent for all the tanks examined, permitting the local sea-level change over the past 2000 years to be established at each location with a precision of ± 20 cm and against which other coastal archaeological features can be calibrated. We conclude that published local sea levels that are based on the present-day elevations of the foundations of protective walls constructed around the tanks and lie ~ 50 cm above our inferred levels are inconsistent with the successful functioning of the water exchange and have to be rejected. In one case, for Santa Liberata, we have been able to calibrate our interpretation against sedimentary evidence from the nearby Orbetello Lagoon that confirm our interpretation of the functional control level of the tanks and we conclude that the accuracy of our local sea levels is also ± 20 cm. The causes of sea-level change along this section of the coast are several, including land motion driven by tectonic and glacio-isostatic processes and any change in ocean volume. The individual estimates for the observed local sea levels range from − 0.9 to − 1.5 m with a mean value of − 1.22 ± 0.20 m. These values show that the spatial variability of the local levels is small and consistent with model-inferences of the glacioisostatic process that indicate near-constant contributions for this section of coast and with tectonic inference from the elevations of the Last Interglacial shoreline.