Prediction of delayed mortality of fire-damaged ponderosa pine following prescribed fires in eastern Oregon, USA

Abstract

Prescribed burning is a management tool used to reduce fuel loads in western interior forests. Following a burn, managers need the ability to predict the mortality of individual trees based on easily observed characteristics. Astudy was established in six stands of mixed-age ponderosa pine (Pinus ponderosa Dougl. ex Laws.) with scattered western junipers at the south end of the Blue Mountains near Burns, Oregon, USA. Stands were thinned in either 1994 or 1995. Three treatments, a fall burn, a spring burn, and an unburned control, were randomly assigned to 12-ha experimental units within each stand. Prescribed burns occurred during mid-October of 1997 or mid-June of 1998 and were representative of operational burns, given weather and fuel conditions. Within each experimental unit, six 0.2-ha plots were established to evaluate responses to the burns. Ponderosa pine plot trees (n =3415) alive 1 month after the burns were evaluated and observed for four growing seasons. Nine fire damage and tree morphological variables were evaluated by logistic regression. A five-factor full model and a two-factor reduced model are presented for projecting probability of mortality. Significant variables in the full model included measures of crown, bole, and basal damage.