We examined fire history and forest age distribution in an unmanaged, Picea-dominated landscape in the Paanajärvi wilderness, located in northeastern Fennoscandia and northwest Russia. Maximum tree age was determined in 61 systematically located study plots in an area of about 6600 ha. Fire scars were examined in living and dead trees in the study plots and elsewhere in the study area. Charcoal and pollen analyses of peat were performed on samples from two locations. Fires had been rare in the landscape. Nearly half of the dendrochronologically dated fires occurred in a distinct and short period, from 1859 to 1889, in the northeastern part of the area. This nonrandom occurrence of fires, together with the observed signs of past human influence, suggests an anthropogenic origin for the majority of the fires. The fact that 95% of the study area consisted of forests older than 120 years reflects the end of the occurrence of fires in the 1880s. Pollen analysis from the southwestern part of the study area showed that the site had been dominated by Picea at least during the last millennium. Charcoal analysis from the same site indicated that likely more than 1000 years had elapsed since the last fire. In general, the results suggest that the abundance of old forests, with the oldest trees being approximately 300 years of age, belongs to the natural state of a Picea-dominated landscape.Résumé : Nous avons étudié l'historique des feux et la distribution de l'âge des forêts dans un paysage non aménagé dominé par Picea dans la région sauvage de Paanajärvi située dans le nord-est de la Scandinavie et le nord-ouest de la Russie. L'âge maximum des arbres a été déterminé dans 61 placettes-échantillons systématiquement réparties dans une zone d'environ 6600 ha. Les cicatrices de feu ont été examinées sur les arbres vivants et les arbres morts dans les placetteséchantillons mais aussi ailleurs dans la zone d'étude. Des analyses du charbon de bois et du pollen présents dans la tourbe ont été réalisées sur des échantillons provenant de deux endroits. Les feux ont été rares dans le paysage. Près de la moitié des feux qui ont été datés au moyen de la dendrochronologie sont survenus dans une période de temps pré-cise et assez courte, soit de 1859 à 1889, dans la partie nord-est de la zone. Le fait que ces feux ne soient pas survenus au hasard, ainsi que la présence d'indices de l'influence humaine passée, indiquent que la majorité des feux seraient d'origine humaine. Le fait que 95 % de la zone d'étude soit couverte de forêts âgées de plus de 120 ans correspond à la fin de la période de récurrence des feux dans les années 1880. L'analyse du pollen provenant de la partie sud-ouest de la zone d'étude montre que le site a été dominé par Picea pendant le dernier millénaire. L'analyse du charbon de bois provenant du même endroit montre qu'il s'est probablement écoulé 1000 ans depuis le dernier feu. En général, les résultats indiquent que l'abondance de vieilles forêts, où les plus vieux arbres ont environ 300 ans, correspond à l'état natu...

The fire history and the effect of fires on the long-term (apparent) rate of carbon accumulation (LORCA) in the Patvinsuo mire complex, eastern Finland, were determined through stratigraphic and pollen analyses of 98 peat cores. The peat cores were characterized by a large number of charcoal layers and the age of the basal peat varied between 57 and 10 500 years. Mire fires slowed down the progress of vertical peat accumulation and resulted in great carbon losses. The average LORCA in the Patvinsuo cores was clearly lower, 9.2 ± 1.0 (SE) g m-2 yr-1, compared to the average value for all mire sites in the southern half of Finland, 17.7 ± 0.6 (SE) g m-2 yr-1. The average rate of carbon loss in Patvinsuo mire complex was 9.5 ± 1.0 (SE) g m-2 yr-1 and the mean carbon loss in an individual fire was estimated to be 2.5 kg m-2.

The charcoal and pollen stratigraphy of an annually laminated sediment of a small lake in eastern Finland (63°109N, 30°589E) was examined, and found to cover a period of 1300 years. The region around the lake is remote and nowadays almost unpopulated. The fire interval around the study lake was 85 years, with a range of 60–134 years during the 900-year period before any major human influence, and the corresponding fire rotation time estimate was c. 130 years. About half of the fires were probably mild surficial fires, leaving most Scots pine alive. Only one extensive, catastrophic fire during the 900-year period was evident. The fires maintained the forest succession cycles before human influence. The abundance of smallest charcoal particles and trends in estimated palaeotemperatures showed congruence, suggesting increased burning in large regions during warmer climatic periods. The charcoal influx values increased markedly from c. ad 1600 onwards due to the commencement of slash-and-burn cultivation in the area. The pollen and charcoal evidence suggests a period of extensive slash-and-burn agriculture from c. ad 1720 to the beginning of the twentieth century. The short fire interval during the slash-and-burning period caused major changes in forest structure: spruce decreased considerably, and pine became the dominant tree in the area. The charcoal influx into the lake has increased since the beginning of the present century despite fire suppression and the charcoal particle-size distribution suggests long distance origin for these particles.

Local fire history covering the entire Holocene period at a dry forest site in North Karelia, eastern Finland (ca. 63°07' N, 30°44' E), was reconstructed on the basis of visible charcoal layers from peat deposits of a small mire basin. Seven points studied along a transect a few metres long provided a record of ancient local forest fires that had scarred the margin of the peat deposit. The charcoal layer records indicate a drastic increase in forest fires about 500 years ago compared with the earlier part of the Holocene period. During the past 500 years, human influence has been extensive in the area, and there have been 9 local fires during that period, while during the previous 9500 years there had been only 34 fires. Between the establishment of spruce (Picea abies (L.) Karst.) 6300 calendar years B.P. and the beginning of significant human influence, the site had burned over at a mean interval of 220–260 years. The data indicate a decrease in fire frequency associated with a warm climate between 9000 and 6300 calendar years BP. This suggests that climatic warming does not necessarily result in increased frequency of forest fires.

No abstract