The prospected strong high-latitude warming is projected to cause a northward move of the arctic-boreal treeline. These changes will potentially have significant effects on climate, but their timing and mode are not well understood. Besides being governed directly by temperature increases, speed of treeline changes will also be influenced by the capacity of migration and establishment, and by competition between different treeline forming species. In Siberia, where about half of the global circum-arctic boreal treeline is located, the treeline is formed by monospecific stands of three species of larch, Larix sibirica, Larix gmelinii and Larix cajanderi. There is clear evidence for ecological separation, particularly between L. sibirica and the northeastern species, with only L. gmelinii and L. cajanderi being able to survive on permafrost with an active layer depth of less than 1-2 m, but L. sibirica being competitively superior at sites with low permafrost tables. Under a warming climate, the respective ranges of the treeline forming species are projected to shift to the Northeast, causing turnovers of forest tree species. The Siberian treeline has undergone several pronounced latitudinal fluctuations in the Holocene, and detailed analyses of the historical processes of vegetation change and competitive displacement would add to predictions for the projected future shift. We are using sedimentary ancient DNA to analyze lake sediment cores spanning most of the Holocene from the southern Taymyr peninsula, where the ranges of L. sibirica and L. gmelinii come together. Changes of the complete vegetation are revealed by DNA metabarcoding and pollen analyses, while diagnostic mitochondrial haplotypes trace the temporal dynamics in distribution of the two closely related larch species. We incorporated these two species into our larch population dynamics model LAVESI to understand the influence competition between these species might have on the speed and timing of treeline movement under changing climates. Simulations were forced with regional climate series at locations in the vicinity of the sampled lakes. These analyses offer a very high degree of resolution and shed light on the complicated ecological processes leading to a change in overall vegetation.