Info

FIGURE 2 Site locations of pollen records along the PEP 1 transect. Locations are given in Table 1.

suspect when we refer to events with a duration of <1000 years. Another problem is related to the dating of carbonates. Radiocarbon dates from lacustrine sediments, at times, can be affected by old carbon (hard-water effect) which produces dates that are too old. Furthermore, the hard-water effect can vary through time in lakes or wetlands. At several coring sites, the dates obtained for different lake levels should be interpreted with great caution, especially in case of isolated ages. In the case of the late glacial, the brief nature of the intervals such as the Older Dryas, IACP, or IBCP requires high temporal resolution, taking into account changes in sedimentation processes that seem to occur in most of the records.

In this chapter, we consider three time intervals corresponding to the Oldest Dryas (15,500-14,500 cal. B.P.), the B0lling-Aller0d interval (14,500-12,700 cal. B.P.), and the Younger Dryas (12,700-11,000 cal. B.P.); all intervals have a duration of > 1000 years, which compares with chronologies obtained from revised sites and laminated sediments (e.g., Cariaco basin). Establishing the presence of the Younger Dryas interval has been complicated by the difficulty of precisely dating its termination with the radiocarbon method. Due to the decreasing concentration of 14C in the atmosphere at that time, radiocarbon dates are nearly the same over an 800-year period (Stuiver and Reimer, 1993).

All radiocarbon dates that cover the late glacial were calibrated (Table 1) by using the CALIB 3.0 program (Stuiver and Reimer, 1993). Late glacial dates older than 10,100 cal. B.P. were calibrated with U/Th and 14C coral sets (Bard et al., 1990, 1993). The 2 sigma standard deviation will be discussed. Only the mean value of the given interval is presented in the text; the whole calibrated interval is given in Table 1. However, we need to keep in mind the probability that a specific age could be the mean value or any other value within this interval. This is an important point to consider when short-term climatic changes such as the late glacial oscillations are discussed. In comparing different vegetation records, therefore, we should consider the entire sequence in order to detect whether there is synchroneity for a specific climatic event or if a temporal lag of climatic events between records is not due to chronological problems. This might be the reason why, in some cases, the interpolated ages do not exactly fit the specific time zone discussed.

We also need to define if in different records there is a difference in the timing of the signal with respect to the beginning or end of the change and if there is a difference in the climatic signal itself. The presence/decrease/increase of the frequency of indicator taxa will be discussed for each record. The presence of the considered indicator taxa will then be translated in terms of warmer/drier/wetter/cooler climatic conditions and compared with the preceding time interval and not with modern data (Figs. 4-6). This climatic reconstruction then will be compared among the records.

0 0

Post a comment