The global human footprint has fundamentally altered wildfire regimes, creating serious consequences for human health, biodiversity, and climate. However, it remains difficult to project how long-term interactions among land use, management, and climate change will affect fire behavior, representing a key knowledge gap for sustainable management. We used expert assessment to combine opinions about past and future fire regimes from 99 wildfire researchers. We asked for quantitative and qualitative assessments of the frequency, type, and implications of fire regime change from the beginning of the Holocene through the year 2300.

Under climate change, modifications on plants’ growth are expected to be the strongest at species margins. Therein, tree acclimation could play a key role as migration is predicted to be too slow to track shifts of bioclimatic envelops. A requirement is, however, that intra-population genetic diversity be high enough for allowing such adaptation of tree populations to climate change. In this study, we tested for the existence of relationships between genetic diversity, site environmental conditions, and the response of annual tree growth to climate of Pinus cembra at its southern limit in the Alps. Site-specific climatic and environmental factors predominantly determined the response of trees along the precipitation gradient. The growth-climate interactions were chiefly linked to mean annual precipitation and temperature, slope and tree-size, and less to genetic diversity. We show that genetic background of Pinus cembra has exclusively indirect modulating power with limited effects on tree-ring formation, and within the southern limit in the Alps, genetic variability is not necessarily well expressed in the patterns of annual tree growth. Our results may imply little adaptive capacity of these populations to future changes in the water balance.

Although lacustrine sedimentary charcoal has long been used to infer paleofires, their quantitative reconstructions require improvements of the calibration of their links with fire regimes (i.e. occurrence, area, and severity) and the taphonomic processes that affect charcoal particles between the production and the deposition in lake sediments. Charcoal particles >150?µm were monitored yearly from 2011 to 2016 using traps submerged in seven head lakes situated in flat-to-rolling boreal forest landscapes in eastern Canada. The burned area was measured, and the above-ground fire severity was assessed using the differentiated normalized burn ratio (dNBR) index, derived from LANDSAT images, and measurements taken within zones radiating 3, 15, and 30?km from the lakes. In order to evaluate potential lag effects in the charcoal record, fire metrics were assessed for the year of recorded charcoal recording (lag 0) and up to 5?years before charcoal deposition (lag 5). A total of 92 variables were generated and sorted using a Random Forest-based methodology. The most explanatory variables for annual charcoal particle presence, expressed as the median surface area, were selected. Results show that, temporally, sedimentary charcoal accurately recorded fire events without a temporal lag; spatially, fires were recorded up to 30?km from the lakes. Selected variables highlighted the importance of burned area and fire severity in explaining lacustrine charcoal. The charcoal influx was thus driven by fire area and severity during the production process. The dispersion process of particles resulted mostly of wind transportation within the regional (<30?km) source area. Overall, charcoal median surface area represents a reliable proxy for reconstructing past burned areas and fire severities.

We used bioproxies from paleosoils buried within two aeolian dunes to test hypotheses concerning the origin of dry sandy boreal forests in Canada. These forests are dominated today by Pinus banksiana Lamb. One hypothesis is that too frequent Holocene stand-replacing fires would have transformed the original vegetation through extirpation of susceptible species to fire in water stress habitat. Alternatively, the ecosystem would have not changed since the dunes stabilized enough to support forest establishment. The vegetation composition and richness were determined by identification of charcoal and macroremains and radiocarbon dating for the chronology. Both sites revealed a similar history covering 6400 years. Half of the charcoal layers were less than 2500 years old in both sites, indicating that they had been subjected to the same fire history. Data indicated a stable plant composition and richness, although the percentage of Pinus decreased slightly over 4000 years (decreasing rate 1% per century). The fungus Cenococcum geophilum was consistently present, with a stochastic abundance. The vegetation grew under natural fire conditions and soil dryness since 6000 years. The ecosystem was probably not stressed by late-Holocene fires or climate changes, as the multi-millennial steady state reveals a resistant and resilient ecosystem.

Genetic processes shape the modern-day distribution of genetic variation within and between populations and can provide important insights into the underlying mechanisms of evolution. The resulting genetic variation is often unequally partitioned within species’ distribution range and especially large differences can manifest at the range limit, where population fragmentation and isolation play a crucial role in species survival. Despite several molecular studies investigating the genetic diversity and differentiation of European Alpine mountain forests, the climatic and demographic constrains which influence the genetic processes are often unknown. Here, we apply non-coding microsatellite markers to evaluate the sporadic peripheral and continuous populations of cembra pine (Pinus cembra L.), a long-lived conifer species that inhabits the subalpine treeline ecotone in the western Alps to investigate how the genetic processes contribute to the modern-day spatial distribution. Moreover, we corroborate our findings with paleoecological records, micro and macro-remains, to infer the species’ possible glacial refugia and expansion scenarios.

In the boreal forest of eastern North America, the distribution of eastern white cedar (Thuja occidentalis) is characterized by a latitudinal fragmentation gradient from south to north. Marginal populations could be outposts allowing cedar to expand its presence in the north in response to climate change. This study aimed to characterize the spatio-temporal regeneration dynamics of 20 marginal cedar stands in order to evaluate their expansion capacity into adjacent black spruce stands. Cedar recruitment within marginal cedar stands was mainly by layering, which allowed the species to maintain for a long time in the landscape. However, the rate of expansion of marginal stands into adjacent black spruce stands through seed dispersal was very low (0.28 m.year?1) and it was negatively influenced by distance to seed trees. Substrate had no significant effect. Global warming could lead to increased seed production by cedar, which could favour the species’ expansion at its northern distribution limit. However, global warming could also increase the frequency and severity of wildfires, which would have a negative effect on cedar expansion capacity.

Tree species responses to climate change will be greatly influenced by their evolutionary potential and their phenotypic plasticity. Investigating tree-rings responses to climate and population genetics at the regional scale is crucial in assessing the tree behavior to climate change. This study combined in situ dendroclimatology and population genetics over a latitudinal gradient and compared the variations between the two at the intra- and inter-population levels. This approach was applied on the northern marginal populations of Thuja occidentalis (eastern white-cedar) in the Canadian boreal forest. We aimed first to assess the radial growth variability (response functional trait) within populations across the gradient and to compare it with the genetic diversity (microsatellites). Second, we investigated the variability in the growth response to climate at the regional scale through the radial growth-climate relationships, and tested its correlation with environmental variables and population genetic structure. Model selection based on the Akaike Information Criteria revealed that the growth synchronicity between pairs of trees of a population covariates with both the genetic diversity of this population and the amount of precipitation (inverse correlations), although these variables only explained a small fraction of the observed variance. At the regional scale, variance partitioning and partial redundancy analysis indicate that the growth response to climate was greatly modulated by stand environmental variables, suggesting predominant plastic variations in growth-response to climate. Combining in situ dendroclimatology and population genetics is a promising way to investigate species' response capacity to climate change in natural stands. We stress the need to control for local climate and site conditions effects on dendroclimatic response to climate to avoid misleading conclusions regarding the associations with genetic variables.

Aim

Towards the cold margins of the Northern Hemisphere boreal zone, continuing warming should theoretically provide a longer vegetative season, favouring growth and a northward shift in tree species distribution. The northern distribution of Thuja occidentalis L. (eastern white cedar) is marked by the presence of isolated marginal populations distant from the continuous distribution. If those populations proved to be well adapted to their future local climatic conditions, their expansion could accelerate cedar poleward migration. We tested the hypotheses that (1) there will be a growth increase in cedar northern marginal populations as a result of global warming, and (2) the edaphic conditions and regional precipitation regimes will modulate their response to warming.

Location

Canadian boreal forest, western Québec (47–50° N, 74–80° W).

Methods

We investigated radial growth using tree-ring measurements from dominant and co-dominant eastern white cedar trees (n = 723) distributed along a latitudinal gradient from the species' northern margin to the centre of its natural range. First, low-frequency growth variations were analysed on whole chronologies (ad 1720–2010). Second, inter-annual growth variations were tested against ad 1953–2010 monthly temperature and precipitation time series with a bootstrapped correlation function. Finally, the impact of environmental variables on the growth–climate relationships was assessed.

Results

Unexpectedly, a growth decline was observed starting in 1980 in marginal sites. Dendroclimatic analyses revealed that radial growth was not only limited by short growing seasons but also by summer droughts in the marginal zone. This response was exacerbated in sites that received less summer precipitation. Counterintuitively, autumn and spring precipitation negatively impacted on growth, especially in wet soil stands.

Main conclusions

Northern marginal populations of cedar may have already reached their optimum temperature threshold for radial growth. Our results suggest that they will probably be facing increasing hydric stress selection pressure under the assumptions of climate change. Their responses to future warming will be highly dependent on the seasonality and magnitude of variation in precipitation regimes.

Paleofire events obtained from the statistical treatment of sedimentary charcoal records rely on a number of assumptions and user's choices, increasing the uncertainty of reconstructio\ns. Among the assumptions made when analyzing charcoal series is the choice of a filtering method for raw Charcoal Accumulation Rate (CHARraw). As there is no ultimate CHARraw filtering method, we propose an ensemble-member approach to reconstruct fire events. We modified the commonly used procedure by including a routine replicating the analysis of a charcoal record using custom smoothing parameters. Dates of robust fire events, uncertainties in fire-return intervals and fire frequencies are derived from members' distributions. An application of the method is used to quantify uncertainties due to data treatment in two CHARraw sequences from two different biomes, subalpine and boreal.

• Strategic introduction of less flammable broadleaf vegetation into landscapes was suggested as a management strategy for decreasing the risk of boreal wildfires projected under climatic change. However, the realization and strength of this offsetting effect in an actual environment remain to be demonstrated.
• Here we combined paleoecological data, global climate models and wildfire modelling to assess regional fire frequency (RegFF, i.e. the number of fires through time) in boreal forests as it relates to tree species composition and climate over millennial time-scales.
• Lacustrine charcoals from northern landscapes of eastern boreal Canada indicate that RegFF during the mid-Holocene (6000–3000 yr ago) was significantly higher than pre-industrial RegFF (ad c. 1750). In southern landscapes, RegFF was not significantly higher than the pre-industrial RegFF in spite of the declining drought severity. The modelling experiment indicates that the high fire risk brought about by a warmer and drier climate in the south during the mid-Holocene was offset by a higher broadleaf component.
• Our data highlight an important function for broadleaf vegetation in determining boreal RegFF in a warmer climate. We estimate that its feedback may be large enough to offset the projected climate change impacts on drought conditions.

In this era of climate change, understanding past and predicting future fire activity are scientific challenges that are central to the development of sustainable forest management practices and policies. Such objectives, however, are difficult to achieve for several reasons. Uncertainties about future fire activity can be superimposed on the short time period covered by existing meteorological data and fire statistics, from which a historical range of variability can be determined. Regional fire activity is also tremendously variable over time, such that contemporary fire records cannot provide information on the full range of fire activity variability a given forest experienced and adapted to. This factor is increasingly important when it comes to determining the resilience of boreal forests to changes in climate and disturbance regimes. In this paper, we present a synthesis of past, present and future trends in seasonal fire danger and fire activity based on data gathered in eastern Canadian boreal forests over the last 20 years, and we provide a critical assessment of the ability to conduct sustainable forest management over the 21st century. The data synthesis provides compelling evidence of a synchronous pattern of decreasing fire-conducive climatic conditions and activity of large fire seasons over the last 2000 years in the eastern coniferous boreal forest. Model simulations suggest that the climate will become drier in upcoming decades, driving future fire activity close to the upper bound of the pre-industrial range of variability. The effects of increasing fire incidence cumulated with forest harvesting may thus pose a risk to forest resilience in the future. This ecological knowledge should help us to define forest management strategies and practices considering future fire activity changes forecasted under climate change. Development of alternative silvicultural interventions that would emulate secondary disturbances (e.g. wind, insects) rather than fire would be necessary to maintain pre-industrial forest characteristics (e.g. composition and age class distribution), and associated forest resilience.

Testate amoebae that inhabit peat are sensitive indicators of water table position. In this study, we used testate amoebae in sediments from a mire in the western Alps (Lac du Thyl) to: (1) reconstruct the hydrology of the site over the last 7,000 years, (2) determine how hydrological changes affected testate amoebae diversity and (3) infer past trophic state shifts. The study site is located in one of the driest valleys of the Alps and is thus very sensitive to hydrological changes. Our study revealed that the water table depth increased (dry conditions) between 5,800 and 4,000 cal year BP. This triggered establishment of a Sphagnum-type peat and acidic conditions from 5,700 to 4,000 cal year BP. These processes were independent of ongoing transformations of the terrestrial vegetation and soil in the catchment area. After 1,690 cal year BP, the depth to the water table decreased (wetter conditions) and a minerotrophic fen developed. At the same time, the diversity of testate amoebae increased, probably as a result of deforestation that supported the expansion of grassland. Climate and land use were apparently more important factors controlling the lake hydrology than were changes in vegetation and soil in the catchment. Testate amoebae diversity was linked to land cover. Changes in pH were controlled indirectly by external forcing (climate), but more directly by fluctuations in the level of the water table (internal forcing) and autogenous expansion of Sphagnum.

The hypothesis that changes in fire frequency control the long-term dynamics of boreal forests is tested on the basis of paleodata. Sites with different wildfire histories at the regional scale should exhibit different vegetation trajectories. Mean fire intervals and vegetation reconstructions are based respectively on sedimentary charcoal and pollen from two small lakes, one in the Mixedwood boreal forests and the second in the Coniferous boreal forests. The pollen-inferred vegetation exhibits different trajectories of boreal forest dynamics after afforestation, whereas mean fire intervals have no significant or a delayed impact on the pollen data, either in terms of diversity or trajectories. These boreal forests appear resilient to changes in fire regimes, although subtle modifications can be highlighted. Vegetation compositions have converged during the last 1200 years with the decrease in mean fire intervals, owing to an increasing abundance of boreal species at the southern site (Mixedwood), whereas changes are less pronounced at the northern site (Coniferous). Although wildfire is a natural property of boreal ecosystems, this study does not support the hypothesis that changes in mean fire intervals are the key process controlling long-term vegetation transformation. Fluctuations in mean fire intervals alone do not explain the historical and current distribution of vegetation, but they may have accelerated the climatic process of borealisation, likely resulting from orbital forcing.

An original method is proposed for estimating past carbon emissions from fires in order to understand long-term changes in the biomass burning that, together with vegetation cover, act on the global carbon cycle and climate. The past carbon release resulting from paleo-fires during the Holocene is examined using a simple linear model between measured carbon emissions from modern fires and sedimentary charcoal records of biomass burning within boreal and cold temperate forests in eastern Canada (Quebec, Ontario). Direct carbon emissions are estimated for each ecozone for the present period and the fire anomaly per kilo annum (ka) v. present day (0 ka) deduced from charcoal series of 46 lakes and peats. Over the postglacial, the Taiga Shield ecozone does not match the pattern of fire history and carbon release of Boreal Shield, Atlantic Maritime, and Mixedwood Plains ecozones. This feature results from different air mass influences and the timing of vegetation dynamics. Our estimations show, first, that the contribution of the Mixedwood Plains and the Atlantic Maritime ecozones on the total carbon emissions by fires remains negligible compared with the Boreal Shield. Second, the Taiga Shield plays a key role by maintaining important carbon emissions, given it is today a lower contributor.

Natural ecosystems have developed within ranges of conditions that can serve as references for setting conservation targets or assessing the current ecological integrity of managed ecosystems. Because of their climate determinism, forest fires are likely to have consequences that could exacerbate biophysical and socioeconomical vulnerabilities in the context of climate change. We evaluated future trends in fire activity under climate change in the eastern Canadian boreal forest and investigated whether these changes were included in the variability observed during the last 7000 years from sedimentary charcoal records from three lakes. Prediction of future annual area burned was made using simulated Monthly Drought Code data collected from an ensemble of 19 global climate model experiments. The increase in burn rate that is predicted for the end of the 21st century (0.45% year^–1 with 95% confidence interval (0.32, 0.59) falls well within the long-term past variability (0.37 to 0.90% year^–1). Although our results suggest that the predicted change in burn rates per se will not move this ecosystem to new conditions, the effects of increasing fire incidence cumulated with current rates of clear-cutting or other low-retention types of harvesting, which still prevail in this region, remain preoccupying.

We present here a 7000-year wildfire reconstruction based on sedimentary charcoal series from five lakes located south of Hudson Bay in eastern boreal North America. The reconstruction shows a significant downward trend in the frequency of large fires from 0.0061 fire·yr⁻¹ ca. 5000 cal yr BP to 0.0033 fire·yr⁻¹ at present. Simulations of fire-season climate based on UK Universities Global Atmospheric Modelling Programme output and reconstructions based on proxy data both indicate a shift toward increasing available moisture in the region between the mid-Holocene and today. We infer that the diminishing trend in wildfire activity was ultimately caused by the steady orbitally driven reduction in summer insolation. Future higher temperatures not compensated for by significant precipitation increases will bring fire frequency back toward its upper limit, recorded between 6000 and 2000 cal yr BP.

We investigated changes in wildfire risk over the 1901−2002 (ad) period with an analysis of broad-scale patterns of July monthly drought code (MDC) variability on 28 forested ecoregions of the North American and Eurasian continents. The MDC is an estimate of the net effect of changes in evapotranspiration and precipitation on cumulative moisture depletion in soils, and is well correlated with annual fire statistics across the circumboreal (explaining 25–61% of the variance in regional area burned). We used linear trend and regime shift analyses to investigate (multi-) decadal changes in MDC and percentage area affected by drought, and kernel function for analysis of temporal changes in the occurrence rates of extreme drought years. Our analyses did not reveal widespread patterns of linear increases in dryness through time as a response to rising Northern Hemisphere land temperatures. Instead, we found heterogeneous patterns of drought severity changes that were inherent to the nonuniformly distributed impacts of climate change on dryness. Notably, significant trends toward increasing summer moisture in southeastern and southwestern boreal Canada were detected. The diminishing wildfire risk in these regions is coherent with widely reported decreases in area burned since about 1850, as reconstructed by dendrochronological dating of forest stands. Conversely, we found evidence for increasing percentage area affected by extreme droughts in Eurasia (+0.57% per decade; P<0.05) and occurrence rates of extreme drought years in Eurasian taiga (centered principally on the Okhotsk–Manchurian taiga, P=0.07). Although not statistically significant, temporal changes in occurrence rates are sufficiently important spatially to be paid further attention. The absence of a linear trend in MDC severity, in conjunction with the presence of an increase in the occurrence rate of extreme drought years, suggest that fire disturbance regimes in the Eurasian taiga could be shifting toward being increasingly pulse dependent.

We investigated changes in wildfire risk over the 19012002 (AD) period with an analysis of broad-scale patterns of July monthly drought code (MDC) variability on 28 forested ecoregions of the North American and Eurasian continents. The MDC is an estimate of the net effect of changes in evapotranspiration and precipitation on cumulative moisture depletion in soils, and is well correlated with annual fire statistics across the circumboreal (explaining 25–61% of the variance in regional area burned).We used linear trend and regime shift analyses to investigate (multi-) decadal changes in MDC and percentage area affected by drought, and kernel function for analysis of temporal changes in the occurrence rates of extreme drought years. Our analyses did not reveal widespread patterns of linear increases in dryness through time as a response to rising Northern Hemisphere land temperatures. Instead, we found heterogeneous patterns of drought severity changes that were inherent to the nonuniformly distributed impacts of climate change on dryness. Notably, significant trends toward increasing summer moisture in southeastern and southwestern boreal Canada were detected. The diminishing wildfire risk in these regions is coherent with widely reported decreases in area burned since about 1850, as reconstructed by dendrochronological dating of forest stands. Conversely, we found evidence for increasing percentage area affected by extreme droughts in Eurasia (10.57% per decade; Po0.05) and occurrence rates of extreme drought years in Eurasian taiga (centered principally on the Okhotsk–Manchurian taiga, P50.07). Although not statistically significant, temporal changes in occurrence rates are sufficiently important spatially to be paid further attention. The absence of a linear trend in MDC severity, in conjunction with the presence of an increase in the occurrence rate of extreme drought years, suggest that fire disturbance regimes in the Eurasian taiga could be shifting toward being increasingly pulse dependent.

The influence of climatic and local nonclimatic factors on the fire regime of the eastern Canadian boreal forest over the last 8000 years is investigated by examining charred particles preserved in four lacustrine deposits. Herein, we compare the distribution of fire-free intervals (FFIs) and the synchronicity of fire events among sites, using Ripley’s K-function to determine the extent of the role of local-scale vs. large-scale processes with respect to fire control. Between 8000 and 5800 cal. BP (calibrated years before present) the climatic and ecological conditions were less conducive to fire events than after this date. After 5800 cal. BP, the number of fires per 1000 years (fire frequency) progressively increased, reaching a maximum ca. 3400 cal. BP. There was a sharp decrease in fire frequency during the last 800 years. Between 8000 and 4000 cal. BP, comparable FFIs and synchronous fire episodes were determined for the study sites. During this period, the fire frequency was predominantly controlled by climate. After 4000 cal. BP, two sites displayed independent fire histories (different FFI distributions or asynchronous fire events), underlining the important influence of local factors, including short-term fuel wetness, characteristics of the watershed and landscape connectivity, in determining fire occurrence. We conclude that climatic changes occurred during the last 4000 years that induced a rise in the water table; this may explain the high spatial heterogeneity in fire history. Current and projected global climatic changes may cause similar spatial variability in fire frequency.

Fire is fundamental to the natural dynamics of the North American boreal forest. It is therefore often suggested that the impacts of anthropogenic disturbances (eg logging) on a managed landscape are attenuated if the patterns and processes created by these events resemble those of natural disturbances (eg fire). To provide forest management guidelines, we investigate the long-term variability in the mean fire interval (MFI) of a boreal landscape in eastern North America, as reconstructed from lacustrine (lake-associated) sedimentary charcoal. We translate the natural variability in MFI into a range of landscape age structures, using a simple modeling approach. Although using the array of possible forest age structures provides managers with some flexibility, an assessment of the current state of the landscape suggests that logging has already caused a shift in the age-class distribution toward a stronger representation of young stands with a concurrent decrease in old-growth stands. Logging is indeed quickly forcing the studied landscape outside of its long-term natural range of variability, implying that substantial changes in management practices are required, if we collectively decide to maintain these fundamental attributes of the boreal forest.

The coniferous boreal forest of northeastern North America is characterized by large and severe fire events and dominated by black spruce (Picea mariana), with scattered patches of balsam fir (Abies balsamea), a species otherwise predominant in the more southern mixedwood boreal forests, characterized by smaller and less severe fire events. Because balsam fir is a late-successional species ill-adapted to fire, this study aimed at determining if the scattered balsam-fir patches found in the coniferous forest were relics of a former fire regime characterized by less frequent and/or severe conflagrations. Fire and vegetation history were assessed for a coniferous forest site through analyses of charcoal, pollen and plant macroremains preserved in lake sediments, peat and hydromorphic forest soil. Pollen and macroremains analyses show that black spruce dominated the local vegetation since deglaciation (c. 8000 cal. yr BP). Balsam fir was abundant around the site during the warm and humid summers of the Hypsithermal (between c. 7000 and 3500 cal. yr BP), before gradually declining during the cool and dry Neoglacial, which was characterized by increased fire frequency and severity. Scattered balsam fir patches in the coniferous forest result from the fragmentation of formerly larger populations and are presently in disequilibrium with climate.© 2008 SAGE publications. All rights reserved.

Fire-made soil erosion should trigger (i) an increase of inorganic sedimentation within lake-basins and (ii) a change of magnetic susceptibility if the burn depth is strong enough to reach the mineral soil and to modify the magnetism of mineral particles. Magnetic susceptibility will also change with the flux of mineral sediments even without a change of their magnetism. Here, we test the role of fire on soil erosion by measuring the mineral accumulation and the magnetic susceptibility in sediments from seven small lakes’ and two dunes’ profiles from East Canada over the Postglacial. Four sites are located in the boreal forest south of James Bay, two in the eastern maritime Quebec and one in the cold temperate south-eastern Ontario. Charcoal accumulation rate is used as a proxy of biomass burning based on the assumption that higher the biomass burning, higher is the charcoal accumulation. The mineral accumulation, deduced from loss-on-ignition residues, is a proxy of erosion process in the lake catchment areas. No relationship is observed between sediment types, sedimentation, magnetic susceptibility and charcoal concentrations in lakes. The patterns of erosion proxies do not match with those of fire, except in dunes. The results suggest that fires have no significant impact on soil erosion in East Canadian forest ecosystems, except in dry-sandy areas. This fact can result from fire severity that is not strong enough to completely burn the humus layer, especially in northern boreal forest characterized by thick soil organic layers. Fire is thus not a significant process affecting the lake sedimentation by soil material input, nor a factor of soil dynamics by rejuvenation of top most soil centimeters over the Postglacial, except in dry sandy areas where dune activity is obviously controlled by burning. r 2006 Elsevier Ltd. All rights reserved.

1 Studies on the variability of natural fire regimes are needed to understand plant responses in a changing environment. Since vegetation changes might follow or trigger changes in fire frequency, climate models suggest that changes in water balance will accompany current global warming, and the response of fire regimes to Holocene hydro-climate changes and vegetation switches may thus serve as a useful analogue for current change. 2 We present high-resolution charcoal records from laminated cores from three small kettle lakes located in mixed-boreal and coniferous-boreal forest. Comparison with some pollen diagrams from the lakes is used to evaluate the role of the local vegetation in the fire history. Fire frequency was reconstructed by measuring the separation of peaks after detrending the charcoal accumulation rate from any background. 3 Several distinct periods of fire regime were detected with fire intervals. Between c. 7000-3000 cal. Year BP, fire intervals were double those in the last 2000 years. Fire frequency changed 1000 years earlier in the coniferous-boreal forest than in the mixed-boreal forest to the south. The absence of changes in combustibility species in the pollen data that could explain the fire frequency transition suggests that the vegetation does not control the long-term fire regime in the boreal forest. 4 Climate appears to be the main process triggering fire. The increased frequency may be the result of more frequent drought due to the increasing influence of cool dry westerly Pacific air-masses from mid to late Holocene, and thus of conditions conducive to ignition and fire spread. In east Canada, this change matches other long-term climate proxies and suggests that a switch in atmospheric circulation 2-3000 years ago triggered a less stable climate with more dry summers. Future warming is moreover likely to reduce fire frequency. 1 Studies on the variability of natural fire regimes are needed to understand plant responses in a changing environment. Since vegetation changes might follow or trigger changes in fire frequency, climate models suggest that changes in water balance will accompany current global warming, and the response of fire regimes to Holocene hydro-climate changes and vegetation switches may thus serve as a useful analogue for current change. 2 We present high-resolution charcoal records from laminated cores from three small kettle lakes located in mixed-boreal and coniferous-boreal forest. Comparison with some pollen diagrams from the lakes is used to evaluate the role of the local vegetation in the fire history. Fire frequency was reconstructed by measuring the separation of peaks after detrending the charcoal accumulation rate from any background. 3 Several distinct periods of fire regime were detected with fire intervals. Between c. 7000-3000 cal. Year BP, fire intervals were double those in the last 2000 years. Fire frequency changed 1000 years earlier in the coniferous-boreal forest than in the mixed-boreal forest to the south. The absence of changes in combustibility species in the pollen data that could explain the fire frequency transition suggests that the vegetation does not control the long-term fire regime in the boreal forest. 4 Climate appears to be the main process triggering fire. The increased frequency may be the result of more frequent drought due to the increasing influence of cool dry westerly Pacific air-masses from mid to late Holocene, and thus of conditions conducive to ignition and fire spread. In east Canada, this change matches other long-term climate proxies and suggests that a switch in atmospheric circulation 2-3000 years ago triggered a less stable climate with more dry summers. Future warming is moreover likely to reduce fire frequency. 1 Studies on the variability of natural fire regimes are needed to understand plant responses in a changing environment. Since vegetation changes might follow or trigger changes in fire frequency, climate models suggest that changes in water balance will accompany current global warming, and the response of fire regimes to Holocene hydro-climate changes and vegetation switches may thus serve as a useful analogue for current change. 2 We present high-resolution charcoal records from laminated cores from three small kettle lakes located in mixed-boreal and coniferous-boreal forest. Comparison with some pollen diagrams from the lakes is used to evaluate the role of the local vegetation in the fire history. Fire frequency was reconstructed by measuring the separation of peaks after detrending the charcoal accumulation rate from any background. 3 Several distinct periods of fire regime were detected with fire intervals. Between c. 7000-3000 cal. Year BP, fire intervals were double those in the last 2000 years. Fire frequency changed 1000 years earlier in the coniferous-boreal forest than in the mixed-boreal forest to the south. The absence of changes in combustibility species in the pollen data that could explain the fire frequency transition suggests that the vegetation does not control the long-term fire regime in the boreal forest. 4 Climate appears to be the main process triggering fire. The increased frequency may be the result of more frequent drought due to the increasing influence of cool dry westerly Pacific air-masses from mid to late Holocene, and thus of conditions conducive to ignition and fire spread. In east Canada, this change matches other long-term climate proxies and suggests that a switch in atmospheric circulation 2-3000 years ago triggered a less stable climate with more dry summers. Future warming is moreover likely to reduce fire frequency.

General circulation model simulations suggest the Earth's climate will be 1-3.5 degreesC warmer by AD 2100. This will influence disturbances such as forest fires, which are important to circumpolar boreal forest dynamics and, hence, the global carbon cycle. Many suggest climate warming will cause increased fire activity and area burned. Here, we use the Canadian Forest Fire Weather Index to simulate future forest fire danger, showing the expected increase in most of Canada but with significant regional variability including a decrease in much of eastern Canada. These results are in general agreement with paleoecological data and general circulation model results from the 6000 calendar years BP interval, which was a time of a warmer climate that may be an analogue for a future climate. ©2001 NRC Canada

Because some consequences of fire resemble the effects of industrial forest harvesting, forest management is often considered as a disturbance having effects similar to those of natural disturbances. Although the analogy between forest management and fire disturbance in boreal ecosystems has some merit, it is important to recognize that it has limitations. First, normal forest rotations truncate the natural forest stand age distribution and eliminate over-mature forests from the landscape. Second, in the boreal mixedwoods, natural forest dynamics following fire may involve a gradual replacement of stands of intolerant broadleaf species by mixedwood and then softwood stands, whereas current silvicultural practices promote successive rotations of similarly composed stands. Third, the large fluctuations observed in fire frequency during the Holocene limit the use of a single fire cycle to characterize natural fire regimes. Short fire cycles generally described for boreal ecosystems do not appear to be universal; rather, shifts between short and long fire cycles have been observed. These shifts imply important changes in forest composition at the landscape and regional levels. All of these factors create a natural variability in forest composition that should be maintained by forest managers concerned with the conservation of biodiversity. One avenue is to develop silvicultural techniques that maintain a spectrum of forest compositions over the landscape.