Holocene ecosystem succession and fire regime evolution at Tannersville Bog, northeastern United States
Abstract. Northern peatlands are among the largest long-term terrestrial carbon reservoirs on Earth, with carbon accumulation jointly regulated by climate, hydrological conditions, and vegetation composition. Future climate warming is expected to accelerate the transition from lakes to peatlands while increasing the likelihood of peatland fires, which can rapidly release carbon accumulated for millennia. However, our understanding of how fire and peatland ecosystem development co-evolve over millennial timescales remains poorly understood. Tannersville Bog (TANB), a temperate boreal peatland in northeastern Pennsylvania, provides an ideal natural laboratory for investigating how northern peatlands are shaped and affected by fire, post-glacial ecosystem succession, and climate. Here, we integrate pyrogenic polycyclic aromatic hydrocarbons (pyroPAHs), n-alkane biomarkers, and sediment composition analyses to reconstruct the Holocene fire and environmental changes of TANB. The combined proxy records identify four major stages of ecosystem succession, from a shallow lake to a rich fen, a fen–bog transitional system, and ultimately a Sphagnum-dominated bog. Correspondingly, multivariate classification of PAH assemblages identifies four distinct fire regimes, which we link to vegetation and peatland-development proxies through ordination analyses. These fire regimes evolved alongside ecosystem succession, reflecting shifts in fuel composition, combustion processes, and fire-source characteristics. The lake-to-peatland transition also reveals that changing depositional and preservation conditions influenced PAH distributions without obscuring major fire-regime transitions. Our results demonstrate that long-term fire-regime evolution was jointly controlled by changes in vegetation composition, peatland hydrology, and basin development, highlighting strong mutual feedback between ecosystem succession and fire over millennial timescales.