Description
Summary:Ongoing climate change is causing and will continue to cause significant changes in the structure and biodiversity of forest ecosystems. Boreal forests, which are among the world's largest forest biomes (32% of the world's forest cover), are particularly affected by such changes. Fires are major perturbators and contribute to the dynamics of these forests. The recent increase in fire activity in these ecosystems raises ecological and socio-economic issues. The consequences of global warming are already being observed in these regions, particularly in Canada and Russia, where warmer winters and springs (2 to 3°C increase) have been recorded since the 1960s. In addition, ongoing climate change is reversing the net carbon storage of boreal forests, which now appears to be shifting from their carbon sink state to a carbon source state. In order to better understand the mechanisms underlying the functioning of this type of forest, and to anticipate future changes, the study of its past dynamics allows us to provide some answers. The aim of this phD is to study and understand the interactions between the dynamics of Holocene fire, vegetation and climate (last 11700 years) in the Republic of Komi, a region of Russia, where these questions have been very poorly documented. Vegetation dynamics during the Holocene were reconstructed from pollen grain analysis and the REVALS model. An open light forest (a particular type of boreal forest) developed in the early Holocene, from 10,000 to 6,000 cal. yr BP. Then, from 10000 to 3500 cal. yr BP, a dark taiga settled. Then from 3500 cal. yr BP, we observe the gradual return of the light taiga. Paleoclimatic reconstructions based on a dual approach, pollen and molecular biomarkers (brGDGts), have shown a climatic optimum between 7000 and 4000 cal. yr BP. The relationship between these results led us to conclude that for the first half of the Holocene, climate had a major influence on vegetation. Then from 3500 cal. yr BP, the change in fire frequency below a threshold value (FRI, ...