Pollen productivity estimates from the forest—tundra ecotone in west-central Sweden: implications for vegetation reconstruction at the limits of the boreal forest

Abstract

Successful reconstruction of vegetation from fossil pollen data in the arctic and alpine tundra, and at the forest—tundra ecotone, is hampered by low pollen production of the major taxa in these environments and the likely dominance of the background pollen component. New approaches to the quantification of vegetation from pollen data (the REVEALS and LOVE models) rely on quantification of background and local pollen by applying models of pollen dispersal and deposition to modern pollen— vegetation survey work. To date, research within this emerging area has focused on lowland environments. This paper presents pollen productivity estimates (PPEs) from the forest—tundra ecotone derived from west central Sweden, discusses the relevant source area of pollen (RSAP) in this environment and estimates the background pollen contribution. The results suggest that the RSAP for moss polsters at the forest—tundra ecotone is c. 500 m, and that PPEs for the major forest tundra ecotone taxa show significant differences compared with previous studies from southern Sweden. Proximity to range limits (controlled by climatic factors) appears to be the main factor causing generally lower pollen production; however, individual taxa respond in different ways, depending on differences in growth behaviour, reproductive strategy and the species and subspecies that represent the pollen taxa. The background pollen component was dominated by arboreal taxa, and contributed c. 60% of the total pollen loading. Graminoids dominated the local component. The PPEs presented here are to date the most appropriate parameters to use for quantification of vegetation at the tundra treeline; however, the data demonstrate that caution needs to be exercised with regards to selection of appropriate PPEs for accurate quantification of taxa at their ecological range.