Long-term forest composition and its drivers in taiga forest in NW Russia

Understanding the processes behind long-term boreal forest dynamics can provide information that assists in predicting future boreal vegetation under changing environmental conditions. Here, we examine Holocene stand-scale vegetation dynamics and its drivers at the western boundary of the Russian taiga forest in NW Russia. Fossil pollen and conifer stomata records from four small hollow sites and two lake sites are used to reconstruct local vegetation dynamics during the Holocene. Variation partitioning is used to assess the relative importance of the potential drivers (temperature, forest fires and growing site wetness) to the long-term stand-scale dynamics in taiga forest. All the main tree taxa, including the boreal keystone species Picea abies (Norway spruce) and Larix sibirica (Siberian larch), have been locally present since 10,000 cal yr bp. The constant Holocene presence of L. sibirica at three small hollow sites suggests a fast postglacial immigration of the species in northern Europe. Picea was present but not dominant at all study sites until its expansion between 8,000 and 7,000 cal yr bp markedly changed the forest structure through the suppression of Betula (birch), Pinus (pine) and Larix. Our results demonstrate that in general, the Holocene forest dynamics in our study region have been driven by temperature, but during short intervals the role of local factors, especially forest fires, has been prominent. The comparison between sites reveals the importance of local factors in stand-scale dynamics in taiga forests. Therefore, the future responses of taiga forest to climate change will be predominantly modulated by the local characteristics at the site.