Reprint of: Arctic planktic foraminiferal assemblages: Implications for subsurface temperature reconstructions

Abstract

Abstract Earlier reconstructions of the inflow of Atlantic Water to Arctic and Subarctic oceans based on foraminiferal proxy data have been obstructed by uncertain quantitative reconstructions of sea-surface and subsurface temperatures. In this study surface sediment samples with undisturbed sediment-water interface from Polar North Atlantic and Barents Sea were retrieved and prepared at the size fractions from 100 μm to 1000 μm. The foraminiferal analyses show that Neogloboquadrina pachyderma constitutes 96–99% of the fauna in Arctic and Polar surface water masses the same result is obtained when investigating the > 150 μm size fraction. However, in Arctic areas influenced by Atlantic Water, additional faunal information is obtained when using the smaller > 100 μm size fraction. In these areas, N. pachyderma is reduced to about 50%, and the relatively small species Turborotalita quinqueloba becomes very frequent. This also applies to the Coastal Water masses, which are dominated by Neogloboquadrina incompta and Globigerinita uvula. Transfer functions using the current dataset based on the > 100 μm size fraction were developed and assessed. Different statistical models were tested, using both seasonal and annual temperature data from 0 m, 10 m, 50 m, and 100 m water depth. The most precise reconstructions of subsurface temperatures were found when using summer temperatures from the 100 m depth level. The transfer function was tested on Holocene foraminiferal records and compared to previous reconstructions. The results show that our new transfer function based on the > 100 μm fraction generally yields lower temperatures at both 10 and 100 m water depth than earlier reconstructions (e.g. Hald et al., 2007). This could be due to the increased number of samples containing both small species and/or the presence of more small specimens representing cold conditions.