Lava Cast Forest - USLCF001 Additional Site Information Karen B. Arabas, Keith S. Hadley, Evan R. Larson Dating Method: Crossdated Sample Storage Location: Department of Environmental and Earth Sciences, Willamette University, Salem, OR. Contact: Dr. Karen Arabas (karabas@willamette.edu) Reference: Arabas, K.B., K.S. Hadley, and E.R. Larson. 2006. Fire history of a naturally fragmented landscape in central Oregon. Canadian Journal of Forest Research 36: 1108-1120. Abstract: We examined the fire history of 11 forest isolates surrounded by lava flows (kipukas) in central Oregon to determine historical differences in fire regimes between kipukas and the surrounding forest, and the role of spatial and environmental variables in fire occurrence. Tree-ring analysis and statistical comparisons show that historical agency records underestimate the number of fires based on the incidence of fire scars. Fires occurred more frequently on kipukas, were typically smaller, and were predominantly lightning-initiated. Except for three widespread fires, fires on kipukas and in the surrounding forest were largely asynchronous. The mean fire-return interval (MFRI) in the surrounding forest decreased following Euro-American settlement and increased on the kipukas with spot-fire removal. This suggests either that forest management and fire exclusion in the surrounding forest decreased fire spread to the kipukas, or that most fires originated on the kipukas. MFRI correlates strongly with distance to the nearest kipuka and a distance-weighted isolation index. The number of fires correlates with elevation change and distance to the nearest kipuka. Fire in naturally fragmented landscapes is influenced by the spatial arrangement of patches, environmental conditions, and human activities. Reconstructing fire histories from forest isolates in the context of their mainland counterparts may have methodological advantages and theoretical implications for forested landscapes characterized by human-imposed insularity. Comments: Files includes 89 cross sections. Each cross section is identified by location (CF = contiguous forest, H = Hoffman Island, K1-K11 = Kipukas 1 through 11) and sample number (rows 3, 4 and 5). Please refer to Arabas et al. 2006 for fire history analysis including sample locations, fire maps, chronologies, seasonality, MFRI, spatial and temporal analysis, and correlation of fire history to environmental variables. Fire History Graphs: Fire History Graphs illustrate specific years when fires occurred and how many trees were scarred. They are available in both PDF and PNG formats. The graphs consist of 2 parts, both of which show the X axis (time line) at the bottom with the earliest year of information on the left and the latest on the right. The Fire Index Plot is the topmost plot, and shows two variables: sample depth (the number of recording trees in each year) as a blue line along the left Y axis, compared with the percent trees scarred shown as gray bars along the right Y axis. Below, the Fire Chronology Plot consists of horizontal lines representing injuries by year on individual sampled trees. Symbols are overlain that denote the years containing the dendrochronologically-dated fire scars or injuries. The sample ID of each tree is displayed to the right of each line. The Composite Axis below represents the composite information from all individual series. The symbols used to represent the fire scars or injuries, and the filters used to determine the composite information, are shown in the legend. These graphs were created using the Fire History Analysis and Exploration System (FHAES). See http://frames.nbii.gov/fhaes/ for more information. The ftp://ftp.ncdc.noaa.gov/pub/data/paleo/firehistory/firescar/northamerica/fire-hist-graphs/uslcf001-mast-chron-graph.pdf file is the master chronology from Arabas et al. 2006.