Temporal changes of the (super 14) C reservoir effect in lakes.

Abstract

Conventional radiocarbon dates for sediment samples from aquatic systems and of coeval terrestrial samples deviate from each other due to the reservoir effect. The reservoir correction is usually assumed to be constant with time for a specific aquatic system. Our studies confirm that seasonal and secular changes are frequent and are governed by the limnological conditions. Lakes have two principal sources of (super 14) C: atmospheric CO (sub 2) and the total dissolved inorganic carbon (TDIC) of the entering groundwater and runoff. The former has values of ca. 100 pMC; the latter usually has a (super 14) C value well below 100 pMC. Atmospheric CO (sub 2) enters the lake by exchange via its surface. The proportions of these two kinds of input determine the magnitude of the reservoir correction in freshwater lakes. It is mainly a function of the volume/surface ratio of the lake and, consequently a function of the water depth. The surface of lakes with outflow does not change when sedimentation decreases the depth of the water. The depth of Schleinsee Lake in southern Germany has decreased from 30 to 15 m since ca. 9000 BP. As a result, the reservoir correction has decreased from ca. -1550 to -580 yr. In contrast, the depth of Lake Proscansko in Croatia increased with growth of the travertine dam and the reservoir correction changed from ca. -1790 to -2650 yr during the last 8800 yr. The largest fluctuations of lake levels occur in closed lakes in arid regions when the climate changes from humid to arid and vice versa. As a result, the reservoir correction of the (super 14) C dates for the total organic fraction from Lejia Lake in the Atacama Desert of Chile varied between <-1800 yr and -4700 yr over a period of only 1800 yr between 11,500 and 9700 BP. The corresponding reservoir correction for the marl fraction is much higher. In summary, accurate and reliable (super 14) C dating of lake sediments requires a study of the temporal changes of the reservoir effect by analysis of both the organic and marl fractions. The most reliable (super 14) C dates are obtained from terrestrial plant remains.