No abstract

We measured apparent marine radiocarbon ages for the Mediterranean Sea, Black Sea, and Red Sea by accelerator mass spectrometry radiocarbon analyses of 26 modern, pre-bomb mollusk shells collected living between AD 1837 and 1950. The marine reservoir (R(t)) ages were estimated at some 390 ± 85 yr BP, 415 ± 90 yr BP and 440 ± 40 yr BP, respectively. R(t) ages in the Mediterranean Sea and Black Sea are comparable to those for the North Atlantic Ocean (<65°N), in accordance with the modern oceanic circulation pattern. The ΔR values of about 35 ± 70 yr and 75 ± 60 yr in the Mediterranean area show that the global box-diffusion carbon model, used to calculate R(t) ages, reproduces the measured marine 14C R(t) ages in these oceanic areas. Nevertheless, high values of standard deviations, larger than measurement uncertainties are obtained and express decadal R(t) changes. Such large standard deviations are indeed related to a decrease of the apparent marine ages of some 220 yr from 1900 AD to 1930 AD in both the Mediterranean Sea and the western North Atlantic Ocean.

S U M M A R YA 37 m long piston core (KC-O1B) was collected from the Calabrian Ridge in the Ionian Sea. Astronomical calibration of sapropels and sapropelic signals-based on rock-magnetic and geochemical properties-yields a very accurate time-frame for the last 1.1 Myr, with the exception of the interval (65&850 ka) straddling the Brunhes/ Matuyama boundary. We used the oxygen isotope data as an additional constraint on our age model.We arrive at an age of 812 ka for the Brunhes/Matuyama boundary, which indicates that its position is determined by delayed acquisition of NRM by some 40 kyr. The age of the top of Jaramillo subchron (1001+3 ka) is slightly older than that of Shackleton, Berger & Peltier ( 1990); that of its bottom ( 1072 f 5 ka) concurs with their estimate.In the Brunhes Chron, four short reversal excursions (CRO-3) were found and dated (CRO 261 f 3 ka, CR1 318 f 3 ka, CR2 515 f 3 ka, CR3 573 3 ka). In all cases, a highcoercivity reversed component is overlapped to a varying degree with a low-coercivity normal component. Three of the reversal excursions (CR1-3) occur in zones where diagenetic processes were shown to be minimal. The fourth (CRO), in which a very large coercivity overlap occurs, is not well expressed and is found in a zone where some dissolution of magnetic minerals has occurred.The existing literature on reversal excursions is reviewed, and some of the older records with an oxygen isotope or sapropel record have been redated. CRO could correspond to one of the Fram Strait excursions (Nowaczyk et at. 1994). CR3 corresponds to the Emperor, which is shown to be equivalent to the Big Lost reversal excursion. For CR1 and CR2 we propose the names Calabrian Ridge 1 and Calabrian Ridge 2, since correlation with existing reversal excursions is uncertain. The Blake was not detected because extensive diagenesis has occurred in the corresponding interval. All reversal excursions observed in KC-O1B correspond to periods of minimal (relative) palaeointensity in the independently dated record of Valet & Meynadier (1993). This seems to imply that reversal excursions are more likely to occur in periods with a relatively large non-dipole field contribution