Abstract. Knowing the historical variation in fire regimes is instrumental in managing forests today and in predicting what may happen in the future. By cross-dating 745 fire scars in 378 samples of remnant Scots pines, we delineated 254 individual forest fires during the past 700 years in a 74-km 2 section of Trillemarka-Rollagsfjell Nature Reserve in south-central Norway. Fire sizes, numbers, burn rates, and frequencies were compared with historical climate proxies, vegetation maps, and written sources. The results revealed patterns consistent with a predominantly climate-driven fire regime up to 1625, followed by periods of strong anthropogenic influence that increased fire frequency during 1600-1700s and diminished fires during 1800-1900s. This was documented by an abrupt increase in number of small fires from the early 1600s that markedly shortened fire intervals from a median of 73 to 37 yr. This shift in fire frequency coincided with a sudden appearance of early-season fires from 1625 and onward. Whereas late-season burn rate increased with summer temperature, no such relationship was found for early-season fires. These results were corroborated by written sources that describe anthropogenic forest fires and slash-and-burn cultivation expanding with the increasing population from the late 1500s and subsequently diminishing due to increasing timber values during 1700-1800s. Whereas human activity strongly influenced the fire regime at multidecadal to centennial scales, it was the interannual variability in climate that triggered large fire events, especially during the pre-1625 period. Prior to 1625, the percentage of years with fire tripled from 7% during cold summers (10-12°C) to 21% during warm summers (14-16°C). Burn rate increased even more, from 0.01% to 1.3% for the same temperature intervals. Ecologically, the post-1625 period is remarkable in such a way that human activity, first by greatly increasing fire frequency and subsequently almost eradicating fires, possibly influenced the fire regime to such an extent that it may be unprecedented for millennia.

This study combines tree-ring and charcoal data to explore possible drivers of the charcoal record and its spatial variation in a boreal Norwegian forest landscape. Peat and mineral soil samples were collected in a multiple site sampling approach and the amount of charcoal in the peat is related to fire history, Holocene climate variation, major shifts in the vegetation composition, and fuel availability. Dendrochronologic dating was used to reveal the fire history over the last 600 years with spatial and temporal accuracy, and AMS radiocarbon dating of 20 peat columns and their charcoal records from four peatlands was used to elucidate the fire history over the Holocene. The average amount of charcoal was about 2.5 times higher in the mineral soil than in the peat (270 versus 100 g/m2, respectively), and there were considerable between- and within-site variations. There was no relationship between the age of a given peatland and its content of charcoal, nor between the amount of charcoal in a given peatland and in the neighboring mineral soil. Although most of the charcoal mass in the peatlands was found in parts of the peat columns originating from relatively warm climatic periods and from the period before the local establishment of Norway spruce ( Picea abies), charcoal accumulation rates (per 1000 yr) were higher during cold climatic periods and similar before and after spruce establishment. Recent fires showed up to a low degree in the peat columns. On fine spatial scales (1–10 m), fuel quality and distribution together with fire behaviour throughout millennia are likely to be responsible for variations in the charcoal record. On the landscape scale (100–1000 m), the charcoal records were site-specifically idiosyncratic, presumably due to topography, distribution of fire breaks and fuel types, and human land use, coupled with long-term variations inherent in these factors.

No abstract

The aim of this study was to ascertain whether there is a relationship between plant species richness and plant-available N, P and water in an environment subject to little anthropogenic disturbance. To accomplish this we studied the vegetation in matorral shrub-lands in northern Patagonia, Argentina. Due to the variation in slope, precipitation and aspect between the sites water status was determined using the 12 C/ 13 C fraction, δ 13 C, to investigate whether this was a confounding factor. The numbers of herb, shrub, liana and tree species were determined at 20 sites along an estimated precipitation gradient. Leaf P and N content and the δ 13 C of Berberis buxifolia were determined, as well as the soil P and N content at the different sites. A negative correlation was found between species richness and Berberis buxifolia foliar P concentration (52% of the species richness variation was accounted for), and a positive correlation was found between plant species richness and Berberis buxifolia foliar N: P ratios (54% of the species richness variation was accounted for). The relationship between species richness and foliar P was seen when all layers of vegetation were included (trees, lianas, shrubs and herbs). Foliar N showed no correlation with species richness, while soil extractable NH 4 showed a weak positive correlation with the number of shrub layer species (lianas, shrubs and trees). The species richness of the shrub layer increased with decreasing values of δ 13 C. Low soil P availability thus affects local species richness in the matorral shrub-lands of Patagonia in Argentina although the growth of vegetation in the area has been shown to be limited by N. We suggest that low P levels increase plant species richness because low soil P concentration is associated with a high P partitioning and high potential for niche separation.