Reconstruction Of Fire History Research Articles

Fire ecology north and south of the Alps since the last ice age

Wildfires are very rare in central Europe, which is probably why fire effects on vegetation have been neglected by most central European ecologists and palaeoecologists. Presently, reconstructions of fire history and fire ecology are almost absent. We analysed sediment cores from lakes on the Swiss Plateau (Lobsigensee and Soppensee) for pollen and charcoal to investigate the relationship between vegetation and fire. Microscopic charcoal evidence suggests increasing regional fire frequencies during the Neolithic (7350-4150 cal. BP, 5400-2200 BC) and the subsequent prehistoric epochs at Lobsigensee, whereas at Soppensee burnings remained rather rare until modern times. Neolithic peaks of charcoal at 6200 and 5500 cal. BP (4250 and 3550 BC) coincided with declines of pollen of fire-sensitive taxa at both sites (e.g., Ulmus, Tilia, Hedera, Fagus), suggesting synchronous vegetational responses to fire at regional scales. However, correlation analysis between charcoal and pollen for the period 6600-4400 cal. BP (4650-2650 BC) revealed no significant link between fire and vegetation at Soppensee, whereas at Lobsigensee increases of Corylus and decreases of Fagus were related to fire events. Fire impact on vegetation increased during the subsequent epochs at both sites. Correlation analyses of charcoal and pollen data for the period 4250-1150 cal. BP (2300 BC -AD 800) suggest that fires were intentionally set to disrupt forests and to provide open areas for arable and pastoral farming (e.g., significant positive correlations between charcoal and Cerealia, Plantago lanceolata, Asteroideae). These results are compared with southern European records (Lago di Origlio, Lago di Muzzano), which are situated in particularly fire-prone environments. After the Mesolithic period (I1 200-7350 cal. BP, 9250-5400 BC), charcoal influx was higher by an order of magnitude in the south, suggesting more frequent fires. Neolithic fires caused similar though more pronounced responses of vegetation in the south (e.g., expansions of Corylus). Post-Neolithic land-use practices involving (controlled) burning culminated in both regions at about 2550 cal. BP (c. 600 BC). However, fire-caused disappearances of entire forest communities were confined to the southern sites. Such differences in fire effects among the sites are explained by the dissimilar importance of fire as a result of different climatic conditions and cultural activities. Our results imply that the remaining (fire-sensitive) fragments of central European vegetation north of the Alps are especially endangered by increasing fire frequencies resulting from predicted climatic change.

Anomalies in grasstree fire history reconstructions for south‐western Australian vegetation

Abstract A new fire history for south‐western Australian sclerophyll forests was proposed recently based on grasstree (Xanthorrhoea preissii ) records that were interpreted to show a high frequency (3–5 years) ‘pre‐European burning regime’. Such a fire regime appears incompatible with the long‐term survival of many fire‐killed woody taxa. We investigated the local fire history in a small area of the northern sand‐plain shrub‐lands of south‐western Australia using 15 grasstrees, examining individual grasstree records in detail and comparing this with the decadal or averaged approach used in the original research, and with fire histories reconstructed from satellite images for the period since 1975. Results lead us to question the utility of the proposed grasstree fire history record as a tool for understanding past fire regimes for two reasons: First, inconsistencies in fire histories among individual grasstrees were considerable – some individuals were not burnt by known fires, while some apparently were burned many times during periods when others were not burned at all. Second, the grasstree record indicates a possible increase in patchiness of fires since 1930, while contemporary evidence and interpretations of the nature of Aboriginal (pre‐European) fire regimes would suggest the opposite. We believe that further research is needed to identify to what extent the grasstree method for reconstruction of fire histories can be used to re‐interpret how fire operated in many highly diverse ecosystems prior to European settlement of Australia.

Reconstruction of the long-term fire history of an old-growth deciduous forest in Southern Québec, Canada, from charred wood in mineral soils

Charcoal particles are widespread in terrestrial and lake environments of the northern temperate and boreal biomes where they are used to reconstruct past fire events and regimes. In this study, we used botanically identified and radiocarbon-dated charcoal macrofossils in mineral soils as a paleoecological tool to reconstruct past fire activity at the stand scale. Charcoal macrofossils buried in podzolic soils by tree uprooting were analyzed to reconstruct the long-term fire history of an old-growth deciduous forest in southern Québec. Charcoal fragments were sampled from the uppermost mineral soil horizons and identified based on anatomical characters. Spruce ( Picea spp.) fragments dominated the charcoal assemblage, along with relatively abundant wood fragments of sugar maple ( Acer saccharum) and birch ( Betula spp.), and rare fragments of pine ( Pinus cf. strobus) and white cedar ( Thuja canadensis). AMS radiocarbon dates from 16 charcoal fragments indicated that forest fires were widespread during the early Holocene, whereas no fires were recorded from the mid-Holocene to present. The paucity of charcoal data during this period, however, does not preclude that a fire event of lower severity may have occurred. At least eight forest fires occurred at the study site between 10,400 and 6300 cal yr B.P., with a dominance of burned conifer trees between 10,400 and 9000 cal yr B.P. and burned conifer and deciduous trees between 9000 and 6300 cal yr B.P. Based on the charcoal record, the climate at the study site was relatively dry during the early Holocene, and more humid from 6300 cal yr B.P. to present. However, it is also possible that the predominance of conifer trees in the charcoal record between 10,400 and 6300 cal yr B.P. created propitious conditions for fire spreading. The charcoal record supports inferences based on pollen influx data (Labelle, C., Richard, P.J.H. 1981. Végétation tardiglaciaire et postglaciaire au sud-est du Parc des Laurentides, Québec. Géographie Physique et Quaternaire 35, 345-359) of the early arrival of spruce and sugar maple in the study area shortly after deglaciation. We conclude that macroscopic charcoal analysis of mineral soils subjected to disturbance by tree uprooting may be a useful paleoecological tool to reconstruct long-term forest fire history at the stand scale.

Fire history and the global carbon budget: a 1°× 1° fire history reconstruction for the 20th century

AbstractA yearly global fire history is a prerequisite for quantifying the contribution of previous fires to the past and present global carbon budget. Vegetation fires can have both direct (combustion) and long‐term indirect effects on the carbon cycle. Every fire influences the ecosystem carbon budget for many years, as a consequence of internal reorganization, decomposition of dead biomass, and regrowth. We used a two‐step process to estimate these effects. First we synthesized the available data available for the 1980s or 1990s to produce a global fire map. For regions with no data, we developed estimates based on vegetation type and history. Second, we then worked backwards to reconstruct the fire history. This reconstruction was based on published data when available. Where it was not, we extrapolated from land use practices, qualitative reports and local studies, such as tree ring analysis. The resulting product is intended as a first approximation for questions about consequences of historical changes in fire for the global carbon budget. We estimate that an average of 608 Mha yr−1burned (not including agricultural fires) at the end of the 20th century. 86% of this occurred in tropical savannas. Fires in forests with higher carbon stocks consumed 70.7 Mha yr−1at the beginning of the century, mostly in the boreal and temperate forests of the Northern Hemisphere. This decreased to 15.2 Mha yr−1in the 1960s as a consequence of fire suppression policies and the development of efficient fire fighting equipment. Since then, fires in temperate and boreal forests have decreased to 11.2 Mha yr−1. At the same time, burned areas increased exponentially in tropical forests, reaching 54 Mha yr−1in the 1990s, reflecting the use of fire in deforestation for expansion of agriculture. There is some evidence for an increase in area burned in temperate and boreal forests in the closing years of the 20th century.

Comparison of charcoal and tree-ring records of recent fires in the eastern Klamath Mountains, California, USA

Fire-history reconstructions are based on tree-ring records that span the last few centuries and charcoal data from lake-sediment cores that extend back several thousand years. The two approaches have unique strengths and weaknesses in their ability to depict past fire events and fire regimes, and most comparisons of these datasets in western conifer forests have focused on sites characterized by high-severity crown fires. Tree-ring and charcoal data spanning the last 300 years in four watersheds in the montane forests of the Klamath Mountains provided an opportunity to compare the records in a fire regime of frequent low- to moderate-severity surface events. The charcoal data were obtained from small lakes, and tree-ring records were derived from fire-scar chronologies at multiple sites within each watershed. The comparison indicates that the tree-ring records detected individual fires not evident in the lake-sediment profiles, whereas the charcoal data disclosed variations in fuel loading and general levels of burning at broader spatial scales. Regional burning in the late 19th and early 20th centuries was evident in the lake-sediment records, and both datasets registered a decline in fire activity in the late 20th century. Thus, the two types of data provide complementary as well as supplementary information on past fire conditions.

Assessing the dendrochronological potential of Pinus occidentalis Swartz in the Cordillera Central of the Dominican Republic

Opportunities for dendrochronology in the tropics are rare because few locations experience sesonality strong enough to induce annual (or even dependably cyclic) ring formation in trees. Consequently, any such opportunities that exist are valuable and identifying them is worth investing time and effort. West Indian pine (Pinus occidentalis Swartz) is endemic to the island of Hispaniola (l9°N, 71°W). It has prviously been shown not to produce useful rings in low-elevation (<2000 m) stands. We investigated its potential in small (4-11 tree) sample sets from four sites between 2000m and 3000m where seasonality of temperature and precipitation is more pronounced. Primary growth rings were distinct at all sites. False rings, diffuse rings, pinching rings and micro-rings were also present, and their number varied considerably by site. Quality of ring formation appeared to increase with both moisture stress (elevation, exposure, edaphic factors) and cold stress (elevation, exposure, cold-air drainage). We were able to cross-date saples (average interseries correlation= 0.44; average mean sensitivity= 0.44) from nine trees at the higheselevation site, and that chronology correlated significantly (r=–0.45, P<0.001) with monthly and sesonal maximum temperatures from a nearby meteorological station. This finding suggests a centuries-long reconstruction of temperature may be possible for this subtropical location. We conclude that, with careful site and tree selection, oversampling to allow exclusion of problematic specimens and greater use of coplete cross-sections rather than cores, P. occidentalis may allow dendrochronological reconstruction of climate, fire history and other environmental processes that operate on the island of Hispaniola.

Late glacial and Holocene vegetation, climate and fire history inferred from Lagoa Nova in the southeastern Brazilian lowland

The pollen and charcoal record of a 1,500-cm-long lake sediment core from Lagoa Nova allows a reconstruction of vegetation, climate and fire history in the Atlantic lowland of southeastern Brazil. Today the potential vegetation of the study region would be dense and tall semi-deciduous forest, related to a dry season of 4 months and 1,250 mm precipitation per year. The lowermost core section of Lagoa Nova is probably of late Glacial age and contains only few, poorly preserved pollen grains, indicating dry environmental conditions. Pollen preservation was good above 1,005-cm core depth. Extrapolating from four radiocarbon dates, this interval represents about the last 10,000 years. During the early Holocene (until about 8,500 b.p.), the landscape was dominated by savanna of the campo cerrado type (open shrub woodland), with frequent trees of Curatella americana and small areas of gallery forests along the drainage system. Savanna fires, probably natural, were frequent at that time. The palaeovegetation pattern is consistent with a long dry season of about 6 months and annual precipitation lower than today. Then, up to about 7,560 b.p., gallery forests expanded in the valleys, reflecting a shorter dry season of about 5 months and somewhat higher annual rainfall. Fire was less frequent during this wetter period. Between about 7,560 and 6,060 b.p., savanna expanded and gallery forests retreated, indicating a return to drier climatic conditions of probably between 5 and 6 months dry season and lower precipitation. Fire was again frequent, but not as frequent as during the early Holocene. During the mid to late Holocene period between about 6,060 and 2,180 b.p., the valleys were covered by semi-deciduous forest, but on the hills cerrado vegetation continued to grow. The dry season was probably around 5 months and rainfall was higher than in the preceding period. During the following period between about 2,810 and 600 b.p., the open cerrado on the hills changed to closed cerrado, reflecting wetter conditions with a shorter dry season than in the previous period. It was only after about 600 b.p. that dense semi-deciduous forests expanded throughout the study region, indicating the start of modern, wet climatic conditions with an annual dry season of about 4 months. Fire was nearly absent during this period. The reconstructed palaeoenvironment from the Lagoa Nova record is similar to that of Lago do Pires, indicating that changes in past vegetation, climate and fire frequency are regionally consistent in southeastern Brazil.

Comparison of pollen-slide and sieving methods in lacustrine charcoal analyses for local and regional fire history

The charcoal content from laminated lake sediments in Québec, Canada, was estimated from pollen slides and by a sieving method. The resulting charcoal series are compared to estimate the suitability of these two methods to provide a local or regional fire history. The replication of five different charcoal series from the sieving method shows that this method is suitable for fire-history reconstruction. In our laminated sediments, 1cm3 is representative of the charcoal content of the sediment. The large charcoal fragments above 15600 mm2are too scarce, however, to provide a significant charcoal series. Comparison of the sieving charcoal series versus the pollen-slide charcoal-series shows that the two series display a roughly similar pattern. The differences between the two series probably result from the accumulation of small particles that have a regional source area and are transported by air over long distances and from high fragmentation rates due to laboratory treatment. Spectral analysis for the last 2000 years shows that the sieving charcoal series have no significant periodic accumulation rate, whereas the spectral analysis of the pollen-slide charcoal series shows a significant period of about 500 years. Because the charcoal particles from the sieving method are larger than those from the pollen-slide method, which are potentially windborne over long distances, our study suggests that the sieving method series is a proxy of local fire history, whereas the pollen-slide method is more suitable for detecting regional trends in fire history.

The use of mineral magnetism in the reconstruction of fire history: a case study from Lago di Origlio, Swiss Alps

The use of magnetic measurements in the detection of fire signals has been neglected since the work of Rummery et al., (1979), yet considerable developments have been made in the interpretation of magnetic measurements over the last 16 years. This paper presents a study of the fire history of Lago di Origlio in the southern Swiss Alps. The study utilises the technique of mineral magnetism alongside the stratigraphic pollen, spore and charcoal records. Correlation between the various proxy records indicates that a magnetic ‘fire’ record is present within the sediments for the last 4 ka. The magnetic fire record has a distinct mineralogical and magnetic grain size signature that can be recognised against the background sedimentary signal. The results suggest that magnetic measurements may be usefully employed in the reconstruction of fire history. Their application is rapid and non-destructive and the results may provide additional information in relation to the links between catchment fire events and the sedimentary record.

Long‐term post‐fire changes in the northeastern boreal forest of Quebec

Abstract. Natural dynamics in the boreal forest is influenced by disturbances. Fire recurrence affects community development and landscape diversity. Forest development was studied in the northeastern boreal forest of Quebec. The objective was to describe succession following fire and to assess the factors related to the changes in forest composition and structure. The study area is located in northeastern Quebec, 50 km north of Baie‐Comeau. We used the forest inventory data gathered by the Ministère des Ressources naturelles du Québec (MRNQ). In circular plots of 400 m2, the diameter at breast height (DBH) of all stems of tree species greater than 10 cm was recorded and in 40 m2 subplots, stems smaller than 10 cm were measured. A total of 380 plots were sampled in an area of 6000 km2. The fire history reconstruction was done based on historical maps, old aerial photographs and field sampling. A time‐since‐fire class, a deposit type, slope, slope aspect and altitude were attributed to each plot. Each plot was also described according to species richness and size structure characteristics. Traces of recent disturbance were also recorded in each plot. Changes in forest composition were described using ordination analyses (NMDS and CCA) and correlated with the explanatory variables. Two successional pathways were observed in the area and characterized by the early dominance of intolerant hardwood species or Picea mariana. With time elapsed since the last fire, composition converged towards either Picea mariana, Abies balsamea or a mixture of both species and the size structure of the coniferous dominated stands got more irregular. The environmental conditions varied between stands and explained part of the variability in composition. Their effect tended to decrease with increasing time elapsed since fire, as canopy composition was getting more similar. Gaps may be important to control forest dynamics in old successional communities.

Variations in fire frequency and climate over the past 17 000 yr in central Yellowstone National Park

A 17000 yr fire history from Yellowstone National Park demonstrates a strong link between changes in climate and variations in fire frequency on millennial time scales. The fire history reconstruction is based on a detailed charcoal stratigraphy from Cygnet Lake in the rhyolite plateau region. Macroscopic charcoal particles were tallied from contiguous 1 cm samples of a 6.69-m-long core, and the data were converted to charcoal-accumulation rates at evenly spaced time intervals. Intervals of high charcoal-accumulation rates were interpreted as local fire events on the basis of information obtained from modern charcoal-calibration studies in the Yellowstone region. The record indicates that fire frequency was moderate (4 fires/1000 yr) during the late glacial period, reached highest values in the early Holocene (>10 fires/1000 yr), and decreased after 7000 calendar yr B.P. The present fire regime (2–3 fires/1000 yr) was established in the past 2000 yr. The charcoal stratigraphy correlates well with variations in July insolation through time, which suggests that regional climate changes are responsible for the long-term variations in fire frequency. In the early Holocene, summer insolation was near its maximum, which resulted in warmer, effectively drier conditions throughout the northwestern United States. At this time, the fire frequency near Cygnet Lake was at its highest. After 7000 calendar yr B.P., summer insolation decreased to present values, the regional climate became cooler and wetter, and fires were less frequent. The Cygnet Lake record suggests that long-term fire frequencies have varied continuously with climate change, even when the vegetation has remained constant.

Pollen and charcoal in lake sediments compared with historically documented forest fires in southern Switzerland since AD 1920

Charcoal in unlaminated sediments dated by 210Pb was analysed by the pollen-slide and thin-section methods. The results were compared with the number and area of forest fires on different spatial scales in the area around Lago di Origlio as listed in the wildfire database of southern Switzerland since AD 1920. The influx of the number of charcoal particles > 75 µm2 in pollen slides correlates well with the number of annual forest fires recorded within a distance of 20-50 km from the coring site. Hence a size-class distinction or an area measurement by image analysis may not be absolutely necessary for the reconstruction of regional fire history. A regression equation was computed and tested against an independent data set. Its use makes it possible to estimate the charcoal area influx (or concentration) from the particle number influx (or concentration). Local fires within a radius of 2 km around the coring site correlate well with the area influx of charcoal particles estimated by the thin-section method measuring the area of charcoal particles larger than 20 000 µm2 or longer than 50 µm. Pollen percentages and influx values suggest that intensive agriculture and Castanea sativa cultivation were reduced 30-40 years ago, followed by an increase of forest area and a development to more natural woodlands. The traditional Castanea sativa cultivation was characterized by a complete use of the biomass produced, so abandonment of chestnut led to an increasing accumulation of dead biomass, thereby raising the fire risk. On the other hand, the pollen record of the regional vegetation does not show any clear response to the increase of fire frequency during the last three decades in this area.

Fire History and Tree Recruitment in an Uncut New England Forest

Fire and forest histories in a hemlock-pine forest in Vermont have been reconstructed by dating fire scars and reconstructing the age distributions of living and dead trees. The ages of living red pines, white pines, and hemlocks show that most of the present forest germinated after a series of spatially overlapping fires between A.D. 1790 and 1850. The ages of cross-dated, dead red pines indicate that this was the third major recruitment interval for pines in this forest since ca. A.D. 1450. We interpret the fire scar and tree age data as recording ca. 50-yr intervals of increased fire frequency recurring every 100-200 yr in response to accumulating fuel loads that coincide with summer drought. The historical records of fires and tree ages, together with the present fuel load, suggest that the next interval of stand-regenerating fires is now overdue. Our success in cross-dating the remnants of dead red pines as old as the 15th century A.D. holds promise for extending reconstructions of fire, forest, and climate history in other parts of this tree's range.

Postglacial Fire Frequency and its Relation to Long-Term Vegetational and Climatic Changes in Yellowstone Park

The paleoecologic record provides unique insights into the response of communities to environmental perturbations of different duration and intensity. Climate is a primary agent of environmental change and its long-term effect on the vegetation of the Yellowstone/Grand Teton region is revealed in a network of pollen records. Fire frequency is controlled by climate, and as climate changes so too does the importance of fire in shaping spatial patterns of vegetation. The prehistoric record of Yellowstone's Northern Range, for example, shows the response of vegetation to the absence of major fires in the last 150 years (Whitlock et al., 1991; Engstrom et al., 1991, 1994). In longer records spanning the last 14,000 years, periods of frequent fire are suggested by sediments containing high percentages of fire-adapted trees, including lodgepole pine and Douglas-fir, and high amounts of charcoal (Whitlock, 1993; Whitlock and Bartlein, 1993; Whitlock et al., 1995; Millspaugh and Whitlock, 1995). The primary research objective has been to study the vegetational history of the Yellowstone region and examine the sensitivity of vegetation to changes in climate and fire frequency on different time scales. This information is necessary to understand better the relative effects of climate, natural disturbance, and human perturbation in the present and future. Fossil pollen and plant macrofossils from dated-lake sediment cores provide information on past vegetation and climate. The frequency of charcoal particles and other fire indicators in dated lake-sediment cores offer evidence of past fires. Studies of future climate and vegetation in the Yellowstone region were based on climate simulations produced by general circulation models that incorporate a doubling of carbon dioxide and an equilibrium model to project potential range displacement of selected tree taxa. Relations between present distributions of tree taxa and climate were established by the use of response surfaces. The study has been divided into three parts: (1) an investigation of the vegetation history of Yellowstone National Park (YNP), (2) an analysis of charcoal accumulation in lakes following the 1988 fires and a reconstruction of fire history in central YNP on long time scales, and (3) a study of the potential ranges of tree taxa in the future with global increases in atmospheric carbon dioxide. A final report to the UW-NPS Research Center describes the methodology and findings in detail (Whitlock et al., 1994).

The reconstruction of boreal forest fire history from lake sediments: A comparison of charcoal, pollen, sedimentological, and geochemical indices

Fossil charcoal, fossil pollen, sedimentological and geochemical analyses of lake sediments have been used previously to reconstruct a history of local fires and resulting vegetation change. The rationale behind these approaches is described and the usefulness of each technique for reconstructing fire history in the boreal forest is assessed empirically. Historical and dendrochronological records provide regional and local fire histories for a site in Wood Buffalo National Park, Alberta, Canada. The local and regional history of fires is compared with the microscopic charcoal content, macroscopic charcoal content, elemental carbon content, fossil pollen content, sedimentology, and geochemistry of annually laminated sediments from a small lake. There is no significant correlation between the abundance of microscopic charcoal, macroscopic charcoal and total elemental carbon content of the sediments. Automated measures of microscopic charcoal abundance made with an image analysis system are correlated significantly with optical counts of microscopic charcoal. None of the charcoal measures provide unequivocal records of local fire activity and the abundance of microscopic charcoal appears to be influenced by variations in regional fire activity. However, the highest depositional rate of macrofossil charcoal occurred during the time of a fire that burned within the drainage basin. Variations in sedimentological measures and geochemistry do not correlate with local fire activity. Fossil pollen percentages and accumulation rates display a pattern of variation that is consistent with observed vegetation responses to fire in the boreal forest. What is likely apparent in the pollen record are the results of a series of burns of sufficient extent and intensity to kill most of the above-ground biomass of the vegetation in an area at least as great as the drainage basin. The inability of evidence from lake sediments to provide detailed histories of past fire activity is not surprising given the wide range of variation in the spatial extent, proximity, intensity and impact of individual fires.

Historical Perspective on the Yellowstone Fires of 1988

Journal Article Historical Perspective on the Yellowstone Fires of 1988: A reconstruction of prehistoric fire history reveals that comparable fires occurred in the early 1700s Get access William H. Romme, William H. Romme Search for other works by this author on: Oxford Academic Google Scholar Don G. Despain Don G. Despain Search for other works by this author on: Oxford Academic Google Scholar BioScience, Volume 39, Issue 10, November 1989, Pages 695–699, https://doi.org/10.2307/1311000 Published: 01 November 1989

The pathogenesis of foot-and-mouth disease

The long-term history of fire regimes in the Mediterranean Andes of Chile is almost unknown. Subalpine woodlands of Austrocedrus chilensis include long-lived trees resilient to low-intensity fires, which can provide valuable tree-ring-based information about fire history. In this work, we performed an annually resolved multicentury reconstruction of past fires from fire-scar records identified in relict Austrocedrus wood found on steep highly-eroded and rocky slopes with coarser fuel structure in the upper Cachapoal Valley, central Andes of Chile. We compared this fire record with historical land-use changes and extensive reconstructions of regional precipitation and temperature, as well as large-scale climatic patterns. The highest fire frequencies were recorded in the Spanish settlement period (1541–1750), when land-clearing activities, cattle ranching, agriculture, and mining practices became widespread after the Spanish conquest. At an interannual time scale, fire occurrence and precipitation were unconnected during the Spanish settlement. By contrast, in the indigenous period (1200–1540), under the influence of the aboriginal Chiquillanes people, fires occurred in wet years with high vegetation productivity. In the livestock grazing period (1751–1950), when large cattle ranches were established, fires occurred in dry years after a wet year. Fires in this period were likely ignited under conditions of high fuel flammability to improve plant production and promote intensive livestock grazing. At a multidecadal time scale, fires were more frequent in cold periods throughout the whole record. These findings suggest that herbaceous fuel accumulation and flammability, modulated by climate variation and human land uses, were the main factors promoting fires spread in this Mediterranean subalpine area. Our research emphasizes the importance of relict Austrocedrus wood for fire history reconstruction and expands knowledge about fire regime shifts over the past eight centuries in southern South America.