Records of past climate variability and associated vegetation response exist in various regions throughout Central and Eastern Europe (CEE). To date, there has been no coherent synthesis of the existing palaeo-records. During an INTIMATE meeting (Cluj Napoca, Romania) focused on identifying CEE paleo-records, it was decided to address this gap by presenting the palaeo-community with a compilation of high-quality climatic and vegetation records for the past 60-8 kyrs. The compilation should also serve as a reference point for the use in the modelling community working towards the INTIMATE project goals, and in data-model inter-comparison studies. This paper is therefore a compilation of up to date, best available quantitative and semi-quantitative records of past climate and biotic response from CEE covering this period. It first presents the proxy and archive used. Speleothems and loess mainly provide the evidences available for the 60-20 ka interval, whereas pollen records provide the main source of information for the Lateglacial and Holocene. It then examines the temporal and spatial patterns of climate variability inferred from different proxies, the temporal and spatial magnitude of the vegetation responses inferred from pollen records and highlights differences and similarities between proxies and sub-regions and the possible mechanisms behind this variability. Finally, it identifies weakness in the proxies and archives and their geographical distribution. This exercise also provides an opportunity to reflect on the status of research in the area and to identify future critical areas and subjects of research.

A) B) Supplementary Figure 1. Photographs of the studied crab shells showing the diversity of color hues and their patterns. Morphological regions from which samples were taken are indicated as blue, red and white areas on Callinectes sapidus (the Atlantic blue crab) claws (A), and green area on Carcinus aestuarii (the Mediterranean green crab) claws (B).

Aim We used fossil records to explore patterns of change in vegetation composition, turnover and diversity along an elevational gradient during the late-glacial to early Holocene, and to locate the elevations most sensitive to past climate changes.Location Romania.Methods Changes in the late-glacial vegetation communities were inferred from seven published pollen records distributed within the main vegetation belts of the Romanian Carpathians, at elevations from 275 to 1840 m. Principal components analysis, detrended canonical correspondence analysis (DCCA) and rarefaction analysis were undertaken on these data.Results DCCA indicates that compositional change is strongest (SD 1.2, c. 70%) at the late-glacial/Holocene transition (c. 11,500 cal. yr bp), but significant shifts also occur at c. 14,700, c. 13,800 and c. 12,700 cal. yr bp (SD 0.4-0.8, 25-50%). Palynological turnover is greater for mid-elevation records (730-1100 m) than at low and high elevations. Intervals of greater palynological richness occur between c. 13,800 and 12,500 cal. yr bp and after 11,500 cal. yr bp, and intervals of lower richness occur before c. 14,000 cal. yr bp and between c. 12,900 and 11,500 cal. yr bp.Main conclusions Variations in species composition during repeated climate changes of the late-glacial suggest that community composition at a given time was not only a result of the environmental conditions of that period, but also the legacy of previous cumulative recruitment and extirpation events. Turnover estimates suggest that mid-elevations have been the most sensitive to climate change during the late-glacial and early Holocene. Palynological richness estimates show a less clear elevational pattern and no evidence for a greater sensitivity of this measure of biodiversity at high elevations to past climate change. However, results may have been affected by taxa with high pollen productivity and distance dispersability. Our finding concurs with other palaeoecological and local-scale modelling studies in suggesting that small populations have survived in favourable microhabitats embedded within larger unsuitable areas during the late-glacial, features not captured by broad-scale model predictions.