PARENT SESSION
Oral Session # 6: Fire Ecology I: Structure and Demographics
Presiding: R Parmenter
Monday, August 4. 8:00 AM to 11:30 AM, SITCC Meeting Room 105.

The event-area relationship: Scale dependence in the fire regime of a New Mexico ponderosa pine forest.

Falk, Donald^*,1, ¹ Ecology & Evolutionary Biology, Tucson, AZ, USA

ABSTRACT- The fire regime is a multivariate characterization of events in space and time. Scale dependence of the fire regime is observed by examining the behavior of metrics of central tendency and higher moments in the distribution of fire intervals. We apply analytical tests adapted from the species-area framework to evaluate the form and underlying mechanisms of scale dependence in the fire regime, substituting fire years for species in a study of an old-growth ponderosa pine forest in the Jemez Mountains of northern New Mexico, USA. Accumulation functions and fire- (species-) richness functions followed the predicted forms, indicating sensitivity to both sample size and area; the latter is the event-area relationship for the number of disturbance events as a function of area. The inverse of frequency, fire interval, is correspondingly scale dependent; mean fire intervals varied by an order of magnitude (2.5-25 yr) for sample areas of 0.5-250 ha. The interval-area function (IAF) is linear in logarithmic space, and the slope of the IAF provides an index of spatio-temporal synchrony. Differences in synchrony of fires are attributable to climatic phases; thus, the slope of the IAF offers an index of the strength of climate entrainment. Probability models of the observed fire interval frequency distributions are positively skewed at small scales (1-50 ha). From first principles in fire ecology, we justify the lognormal distribution as the default model for fire intervals in a surface fire regime, considering fires as multiplicative processes involving a large number of contingent factors. Each of these contributory factors has its own area-scaling function, producing scale-dependence in the fire interval probability distribution as a whole. Scale dependence in the parameters of probability distributions generates the observed sensitivity in metrics of the fire regime to area, creating a synthesis between empirical observations and underlying ecological and statistical theory.

Key words: fire intervals, probability models, scale dependence, event-area relationship