Palaeomagnetism was the method used for dating sediments older than the time span covered by AMS 14C dating. Geomagnetic palaeointensities recorded in Lake Baikal sediments were tuned to a reference curve (the record from ODP Site 984, Channell, 1999) whose chronology is well constrained (Demory et al., 2005a-this volume and Demory et al., 2005b-this volume). The palaeointensity record from ODP Site 984 is of high quality, is well dated and covers the time span of the present study. Anchored by a geomagnetic excursion (the Iceland basin event, dated at 186–189 ka according to Channell et al. (1997)), this age model is constrained by 55 correlation points for a time span of ca. 200 ky. The age models for both core sections in the interval 100–150 ky are shown in Fig. 2.
C/N mass ratios remain constant throughout MIS 3 and into MIS 2, with values between 6.3 and 8.9, indicating no significant terrestrial input of organic matter (Fig. 3). Low %TOC values during the interstadial increase from 0.4 to 0.7 between 57.8 and 43.7 kyr BP with a concurrent gradual increase in δ13C(organic) amid oscillations between −23.2‰ and −26.1‰ (Fig. 3). %TOC falls to 0.4 between 40.9 and 39.4 kyr BP whereas δ13C(organic) remains high at c. 24‰ with a peak value of −23.6‰ at 39.4 kyr BP. The subsequent two-stage increase in %TOC from 39 to 37.9 kyr BP and between 37.3 and 36.9 kyr BP is marked by a period of δ13C(organic) lowering to c. −26.6‰ before δ13C(organic) increases after 37.9 kyr BP to −24.8‰, values comparable to those prior to the %TOC decline at 40.9 kyr BP.
All diatoms in the analysed section were extensively affected by dissolution with only c. 1% of valves in a “pristine” condition. Diatom concentrations were generally extremely low throughout MIS 3 and across the MIS 3/2 transition with samples containing a mixture of extant and extinct species (Fig. 4).