Holocene forest-line and climate dynamics in the Rondane-Dovre area, S-Norway, inferred from pine ( Pinus sylvestris ) megafossils and pollen data, with the emphasis on the early Holocene and the Holocene thermal maximum (HTM)

The result of 97 megafossils of pine from the Rondane-Dovre area is here presented. Of these, 84 megafossils are published for the first time. As the regional present forest-lines are situated at varying elevations, the elevations of megafossil finds are adjusted accordingly and thereafter corrected for land uplift before making paleo-ecological inferences. Our findings show that pine established a vertical belt of ca. 250 m around 9.5 ka BP. Around 8–9 ka BP, the pine-forest line reached the Holocene maximum elevations situated ca. 180 m higher than today, indicating that the July mean was warmest in the early Holocene at least 1.2°C warmer than present. Thereafter, pine forests declined and were replaced by the less warmth-demanding birch forests. A comparison with other Fennoscandian paleoecological studies shows that pine immigrated to S-Norway around 10.3 ka BP from the south and the east which is later than the immigration of the more warmth-demanding Hippophaë. This indicates that in Fennoscandia, the Holocene thermal maximum (HTM), at least regarding summer temperatures, was reached in the early Holocene around 9–10 ka BP, a fact that seems overlooked in the ongoing discussion of the timing of HTM. After 8 ka BP, our findings show that colder summers caused decreases in pine pollen production whereas pine-forest-lines were less affected. This shows how temperatures could influence on pollen production and obscure pollen-based reconstructions of past vegetation. During the 8.2 cold event, winter stress prevented pine regeneration. In the period 4.9–1.5 ka BP, the regional pine-forest line descended below the present, probably due to colder climate. Locally, human impact could also have been involved. After 1.5 ka BP, reappearing megafossils could reflect the Medieval warm period. However, data is too scarce to establish elevations of forest-lines.