Master’s Thesis Abstract by Lara-Karena B. Kellogg (2004)

Fire regimes are complex systems that represent an aggregate of spatial and temporal events whose statistical properties are scale-dependent. Despite the breadth of research regarding the spatial controls on fire regime variability, there are few datasets available with sufficient resolution to test spatially explicit hypotheses. I decomposed the spatial relationships within an extensive, spatially distributed network of geo-referenced, fire-scarred trees (17,700 scars) for six sites in eastern Washington. I utilized the spatial autocorrelation in fire history data to derive empirical and theoretical parameter estimates of semivariance that enabled us to infer mechanisms that generate spatial patterns of fire in ecosystems. I used the Mantel's test on time series of fire occurrence to differentiate the spatial component of their variability from the influences of environmental conditions.

The spatial dependence of historical fire regimes varied within and among sites. Spatial controls on low-severity fire regimes within similar dry forest ecosystem types operate at varying spatial scales, reflecting topographic properties of local landscapes. However, only portions of the spatial variability in fire events can be attributed to topography. In complex, rugged terrain, modal fire sizes associated with the effective ranges in variogram models were 150 ha or less, whereas in more open and rolling terrain, the spatial scale of fire occurrence was not controlled by landform. Results illustrate that the statistical spatial characteristics of fire regimes change with landform characteristics within a forest type, suggesting that a simple relationship between fire frequency and forest-type does not exist. Quantifying the spatial structures in fire occurrence associated with topographic variation demonstrated that fire regime variability is scale and location dependent. By identifying the scale dependencies associated with specific fire regimes we can match the regime to the scales of the controlling factors with greater precision, thus increasing our abilities to evaluate their relationship.