(Tables 1-3) Amphipod species observed under Antarctic pack ice during POLARSTERN cruises ANT-XXI/4, ANT-XXII/2 and ANT-XXIII/6

During three Antarctic expeditions (2004, ANT XXI-4 and XXII-2; 2006, ANT XXIII-6) with the German research icebreaker R/V Polarstern, six different amphipod species were recorded under the pack ice of the Weddell Sea and the Lazarev Sea. These cruises covered Austral autumn (April), summer (December) and winter (August) situations, respectively. Five of the amphipod species recorded here belong to the family Eusiridae (Eusirus antarcticus, E. laticarpus, E. microps, E. perdentatus and E. tridentatus), while the last belongs to the Lysianassidea, genus Cheirimedon (cf. femoratus). Sampling was performed by a specially designed under-ice trawl in the Lazarev Sea, whereas in the Weddell Sea sampling was done by scuba divers and deployment of baited traps. In the Weddell Sea, individuals of E. antarcticus and E. tridentatus were repeatedly observed in situ during under-ice dives, and single individuals were even found in the infiltration layer. Also in aquarium observations, individuals of E. antarcticus and E. tridentatus attached themselves readily to sea ice. Feeding experiments on E. antarcticus and E. tridentatus indicated a carnivorous diet. Individuals of the Lysianassoid Cheirimedon were only collected in baited traps there. Repeated conventional zooplankton hauls performed in parallel to this study did not record any of these amphipods from the water column. In the Lazarev Sea, E. microps, E. perdentatus and E. laticarpus were regularly found in under-ice trawls. We discuss the origin and possible sympagic life style of these amphipods.

Meteorological measurements from 8 dropsondes released during POLAR 4 flight on 1993-03-04 along a track orthogonal to the pack ice edge north west of Svalbard

Different parameterizations of subgrid-scale fluxes are utilized in a nonhydrostatic and anelastic mesoscale model to study their influence on simulated Arctic cold air outbreaks. A local closure, a profile closure and two nonlocal closure schemes are applied, including an improved scheme, which is based on other nonlocal closures. It accounts for continuous subgrid-scale fluxes at the top of the surface layer and a continuous Prandtl number with respect to stratification. In the limit of neutral stratification the improved scheme gives eddy diffusivities similar to other parameterizations, whereas for strong unstable stratifications they become much larger and thus turbulent transports are more efficient. It is shown by comparison of model results with observations that the application of simple nonlocal closure schemes results in a more realistic simulation of a convective boundary layer than that of a local or a profile closure scheme. Improvements are due to the nonlocal formulation of the eddy diffusivities and to the inclusion of heat transport, which is independent of local gradients (countergradient transport).

ELPA (European Leaf Physiognomic Approach): Grid data set of the leaf physiognomic composition of the extant European hardwood vegetation

Physiognomic traits of plant leaves such as size, shape or margin are decisively affected by the prevailing environmental conditions of the plant habitat. On the other hand, if a relationship between environment and leaf physiognomy can be shown to exist, vegetation represents a proxy for environmental conditions. This study investigates the relationship between physiognomic traits of leaves from European hardwood vegetation and environmental parameters in order to create a calibration dataset based on high resolution grid cell data. The leaf data are obtained from synthetic chorologic floras, the environmental data comprise climatic and ecologic data. The high resolution of the data allows for a detailed analysis of the spatial dependencies between the investigated parameters. The comparison of environmental parameters and leaf physiognomic characters reveals a clear correlation between temperature related parameters (e.g. mean annual temperature or ground frost frequency) and the expression of leaf characters (e.g. the type of leaf margin or the base of the lamina). Precipitation related parameters (e.g. mean annual precipitation), however, show no correlation with the leaf physiognomic composition of the vegetation. On the basis of these results, transfer functions for several environmental parameters are calculated from the leaf physiognomic composition of the extant vegetation. In a next step, a cluster analysis is applied to the dataset in order to identify "leaf physiognomic communities". Several of these are distinguished, characterised and subsequently used for vegetation classification. Concerning the leaf physiognomic diversity there are precise differences between each of these "leaf physiognomic classes". There is a clear increase of leaf physiognomic diversity with increasing variability of the environmental parameters: Northern vegetation types are characterised by a more or less homogeneous leaf physiognomic composition whereas southern vegetation types like the Mediterranean vegetation show a considerable higher leaf physiognomic diversity. Finally, the transfer functions are used to estimate palaeo-environmental parameters of three fossil European leaf assemblages from Late Oligocene and Middle Miocene. The results are compared with results obtained from other palaeo-environmental reconstructing methods. The estimates based on a direct linear ordination seem to be the most realistic ones, as they are highly consistent with the Coexistence Approach.