Fire and climate change during the last 750 yr in northwestern Minnesota

Charcoal stratigraphic analysis and fire scars on red pine (Pinus resinosa) trees were used to determine spatial and temporal occurrence of fire in 1 km² of old-growth mixed conifer/hardwood forests in northwestern Minnesota. Charcoal was analyzed year by year on petrographic thin sections from annually laminated sediments of three small (≤5 ha) lakes having adjacent catchments. Dated fire scars (n = 150) from recent treefalls provided an independent record of the spatial patterns of past burns. Sedimentology of the varved sediments, water-balance models that use 150 yr of instrumental temperature and precipitation data, and published data were used to identify climate changes in separate studies, and they were used in this study to examine the possible connection between changing fire regimes and climate change. Fire-history data were used to show the changing probability of fire with time since the last fire and the effects of spatial variance (slope and aspect) on the distribution of fires through time. Over the last 750 yr, fire was most frequent (8.6 ± 2.9-yr intervals) during the warm/ dry 15th and 16th centuries. Intervals were longer (13.2 ± 8.0 yr) during cooler/moister times from AD 1240 to 1440 and since 1600 (the Little Ice Age). The fire regime during the Little Ice Age consisted of periods during the mid-18th and mid-19th centuries characterized by longer fire intervals of 24.5 ± 10.4 and 43.6 ± 15.9 yr, respectively, and short-term warm/dry periods from 1770 to 1820 and 1870 to 1920 when intervals were 17.9 ± 10.6 and 12.7 ± 10.1, respectively. The probability of fire increased through time, probably in step with fuel accumulation. South- and west-facing slopes burned more frequently than did north and east aspects. Fire suppression began in 1910. During warm periods, probability of fire was sufficiently high that a continuous litter layer was all that was necessary for fire to spread and scar trees. During cool and moist times fire was most-likely to occur in years with higher moisture deficits. The combined methods for fire-history analysis provided a more detailed spatial and temporal documentation of fire regimes than has previously been possible from analysis of fire scars or of charcoal counts derived from fossil pollen preparations. Results support predictions of particle-motion physics that thin sections record a local fire history. Because climate varies continuously, the responsiveness of disturbance regime to short- and long-term climatic change suggests caution in the interpretation of fire frequencies that derive from space/time analogies or extrapolation from short-term data.

Duke Scholars

Author James S. Clark Environmental Natural Science