MODELING THE DYNAMICS OF SNAGS

Many wildlife species require standing dead trees (i.e., snags) as par t of their habitat. Therefore, the ability to predict future density, distribution, and condition of snags can assist resource managers in m aking land-use decisions. Here we present methods for modeling the dyn amics of snags using data from a 10-yr study on the rates of decay, fa lling, and recruitment of snags on burned and unburned plots in the Si erra Nevada, California. Snags (all species) in advanced stages of dec ay usually fell within 5 yr, and snags created by fire decayed rapidly and fell quicker (within 10 yr) than those on unburned plots. Pine (P inus spp.) snags decayed more rapidly than fir (Abies spp.). Although there was an overall net increase in snag density on unburned plots, m ost of this increase was in the smaller (> 13-38 cm diameter at breast height [dbh]) size classes; there was a net decrease in the larger (> 38 cm dbh) snags preferred by many birds for nesting and feeding. Ove rall, snags remained standing the longest that were larger in diameter , shorter in height, less decayed, fir rather than pine, and lacking t ops. A Leslie matrix model of snag dynamics predicted changes in snag decay and density only when adjusted for the specific environmental fa ctor(s) causing initial tree mortality. Many snags are created by epis odic events, such as fire, disease, drought, and insects. Models of sn ag dynamics must include the species and condition of trees becoming s nags, as well as the factor(s) causing the tree to die. Forest manager s must consider this episodic creation of snags when developing snag-m anagement guidelines, and when planning tree-salvage programs based on short-term inventories.