The chronology of the Last Interglacial relative sea-level in the Mediterranean Basin: New data from Grotta del Diavolo (Southern Italy)

PUBBLICATO: 28 Settembre 2026

Quaternary Science Reviews, Isola et al. (2026) - Fig. 6 — Fig. 6. Last Interglacial (LIG) Relative Sea Level (RSL) data. From the top: Relative Sea Level curves (Brown line Rohling et al.,(2019); dark green line Grant et al. (2014); Relative sea-level data terrestrial limiting point – speleothems – (red: Andrews et al. (2007); light blue: Bini et al. (2020); Isola et al. (2024); black: Pieruccini et al. (2022); green: Rolfo et al. (2023); orange: Giaccio et al. (2024); light green: De Santis et al. (2024, 2026); brown this work); the GIA output from Antonioli et al. (2018): ANICE-SELEN (green line de Boer et al., 2017); ICE-5G red line (Peltier, 2004) and ICE-6G (blue line Peltier et al., 2015). The vertical bar highlights the LIG highstand time interval. Altitudinal error is too low to be highlighted. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Isola I., M. Bini, A. Columbu, V. De Santis, G. Mastronuzzi, G. Scicchitano, C. Hai, X. Jia, C.-Y. Huang, S. Chuan-Chou, G. Zanchetta (2026).
Quaternary Science Reviews, 391. https://doi.org/10.1016/j.quascirev.2026.110230

Abstract

The precise chronology of the Last Interglacial (LIG) highstand marine deposits in the Mediterranean basin is still elusive, and comparison with the glacial isostatic adjustment (GIA) model and global reconstruction remains challenging. In this paper, we revaluated the stratigraphy of the Grotta del Diavolo coastal cave in the Apulia region and we re-dated a suite of speleothems (flowstones) with U/Th methods, which allow us to constrain the main phases of marine ingression into the cave, supplying new insight for the reconstruction of the relative sea level (RSL) curve in the Mediterranean basin. A marine calcarenite layer was related to LIG. This layer is sandwiched by two flowstones, which were chronologically constrained to be younger than ∼130 thousand years ago (ka, relative to AD, 1950) and older than ∼102 ka. Elevation measurement and U/Th dating indicate that RSL during Termination II (TII) was lower than (at least) ∼3 m a.s.l. at ∼130 ka, giving an important limiting point for the TII RSL transgression. These new data, summed with new recently dated speleothems and corals from southern Italy, allow us to estimate the duration of the LIG highstand from <130 ka to >120 ka.

Back to top