No abstract

Fire is a key Earth system process, with 80% of annual fire activity taking place in grassland areas. However, past fire regimes in grassland systems have been difficult to quantify due to challenges in interpreting the charcoal signal in depositional environments. To improve reconstructions of grassland fire regimes, it is essential to assess two key traits: (1) charcoal count, and (2) charcoal shape. In this study, we quantified the number of charcoal pieces in 51 sediment samples of ponds in the Great Plains and tested its relevance as a proxy for the fire regime by examining 13 potential factors influencing charcoal count, including various fire regime components (e.g. the fire frequency, the area burned, and the fire season), vegetation cover and pollen assemblages, and climate variables. We also quantified the width to length (W:L) ratio of charcoal particles, to assess its utility as a proxy of fuel types in grassland environments by direct comparison with vegetation cover and pollen assemblages. Our first conclusion is that charcoal particles produced by grassland fires are smaller than those produced by forest fires. Thus, a mesh size of 120μm as used in forested environments is too large for grassland ecosystems. We recommend counting all charcoal particles over 60μm in grasslands and mixed grass-forest environments to increase the number of samples with useful data. Second, a W:L ratio of 0.5 or smaller appears to be an indicator for fuel types, when vegetation surrounding the site is before composed of at least 40% grassland vegetation. Third, the area burned within 1060m of the depositional environments explained both the count and the area of charcoal particles. Therefore, changes in charcoal count or charcoal area through time indicate a change in area burned. The fire regimes of grassland systems, including both human and climatic influences on fire behavior, can be characterized by long-term charcoal records.

Fire is one of the most prevalent disturbances in the Earth system, and its past characteristics can be reconstructed using charcoal particles preserved in depositional environments. Although researchers know that fires produce charcoal particles, interpretation of the quantity or composition of charcoal particles in terms of fire source remains poorly understood. In this study, we used a unique four-year dataset of charcoal deposited in traps from a native tallgrass prairie in mid-North America to test which environmental factors were linked to charcoal measurements on three spatial scales. We investigated small and large charcoal particles commonly used as a proxy of fire activity at different spatial scales, and charcoal morphotypes representing different types of fuel. We found that small (125-250 μm) and large (250 μm-1 mm) particles of charcoal are well-correlated (Spearman correlation=0.88) and likely reflect the same spatial scale of fire activity in a system with both herbaceous and woody fuels. There was no significant relationship between charcoal pieces and fire parameters <500 m from the traps. Moreover, local area burned (<5 km distance radius from traps) explained the total charcoal amount, and regional burning (200 km radius distance from traps) explained the ratio of non arboreal to total charcoal (NA/T ratio). Charcoal variables, including total charcoal count and NA/T ratio, did not correlate with other fire parameters, vegetation cover, landscape, or climate variables. Thus, in long-term studies that involve fire history reconstructions, total charcoal particles, even of a small size (125-250 μm), could be an indicator of local area burned. Further studies may determine relationships among amount of charcoal recorded, fire intensity, vegetation cover, and climatic parameters.

Summary1. This study investigated the fire-vegetation relationship by reconstructing the long-term fire and vegetation dynamics around a small lake in the Mediterranean montane belt on Corsica Island. The vegetation is characterised by forests dominated by Pinus nigra ssp. laricio, an endemic subspecies that is currently threatened. Populations of this taxon are geographically restricted, and their ranges are decreasing, possibly because of disturbance, that is, fire and logging.2. Here, we examine the role of fire in vegetation dynamics and its effect on diversity since the lateglacial at a montane site using sedimentary plant macrofossils and charcoal to reconstruct local vegetation and fire frequency, respectively.3. Macrofossils of P. laricio provide evidence for its growth around the Lac de Creno since 13 200 cal. years BP, which is consistent with its natural origin in Corsica. The onset of the Holocene (~11 700 cal. years BP) was marked by a rapid shift in the treeline, the establishment of P. laricio-dominated woodlands with low species turnover, and a long-term increase in taxa richness as a result of successive expansions of broadleaved deciduous trees. In spite of fire-return intervals (FRIs) fluctuating between 30 and 490 years during the Holocene, P. laricio-dominated woodlands persisted for several millennia, and fires likely played an ecological role in the functioning of these woodlands. The 12 000 year record of mean FRI (80 years between fires; i.e. frequency of 12.5 fires 1000 year À1 ) can be used to define a baseline for the management and conservation of P. laricio montane forests. Our findings demonstrate that P. laricio has a long history, a natural origin in Corsica, and that Corsican pine forests have survived to a wide range of burning conditions over the last 13 200 years. Synthesis:We present multimillennial fire and vegetation histories of a Mediterranean montane site. Pinus nigra ssp. laricio has been present near the study site since approximately 13 200 cal. years BP, demonstrating that this pine was present on the island prior to the arrival of pre-historic humans. The long-term records also show that P. laricio woodlands were mixed with deciduous broadleaf trees (Betula sp. and Fagus sylvatica), and other needleleaf trees (Abies alba), at least, and were not influenced by changes in fire frequency. We conclude that (i) fire is a natural component of the ecosystem and (ii) fires likely played an important ecological role in the functioning of the Corsican black pine woodland ecosystem.