Forest Investment Account (FIA) - Forest Science Program
FIA Project Y093017

    Rates of mortality and dead tree dynamics in old coastal forest stands
Project lead: Parish, Roberta (Ministry of Forests and Range)
Author: Parish, Roberta
Subject: Forest Investment Account (FIA), British Columbia
Series: Forest Investment Account (FIA) - Forest Science Program
In low elevation, productive forests on the coast of British Columbia, old stands are being retained for non-timber purposes or remaining in the inventory for long periods due to the age structure of the management unit. Mortality of canopy trees is part of the dynamics of old stands, and dead trees are important structural components of such stands. Standing dead trees and downed logs provide a variety of attributes and functions, including wildlife habitat as well as part of the carbon and nutrient cycles.

The causes of mortality vary greatly and death may result from a combination of factors (Maser 1988; Pedersen 1998). Both mortality rates and the transition from snag to downed wood depend on species and on site factors (Schmid and Hinds 1974; Franklin et al. 1987; Raphael and Morrison 1987; Garber et al. 2005). For example, mortality rates are higher in high-productivity sites than in low-productivity sites (Franklin et al. 1987). The agent of mortality can influence snag longevity (Keen 1929; Schmid et al. 1985), which ranges from instantaneous in the case of wind throw of green trees to over 270 years for large Thuja plicata snags (Daniels et al. 1997).

In order to manage forests retained for non-timber purposes or retained for long periods due to the age structure of a management unit, managers need information on both rates of mortality and rates of transition from standing snags to fallen logs. This knowledge is applicable to a range of considerations, from the safety hazard posed by individual trees to the characteristics of wildlife tree patches. Interviews with those developing policy and guidelines for wildlife trees and coarse woody debris retention found that most BC policies were based on information from the literature from the US Pacific Northwest (Stone 1997). Subsequently, a protocol was developed to collect data relevant to the varied conditions in BC ecosystems (Forest Productivity Council 1999). Since that time, data on dead trees have been collected during regular re-measurements of Ministry of Forests and Range growth and yield permanent sample plots (PSPs).

Data from PSPs were analysed to determine fall rates for two coastal species, Douglas-fir and western hemlock. A logistic regression model based on diameter at breast height (dbh) and time since death provided the best fit (Stone et al. 2002). These relationships were subsequently coded into TIPSY (Table Interpolation Program for Stand Yield, Mitchell et al. 2000), a program used by timber supply specialists in every timber supply analysis in BC. Although the TIPSY components were derived from a limited number of PSPs, (46 for Douglas-fir and 16 for western hemlock), numerous operational applications have relied on the resultant estimates. The deficiencies are documented, but few users consider the consequences. As well, PSPs represent mainly young to moderate-aged stands; thus validation of extrapolations to old stands is problematic.

To augment current information, we plan to combine two approaches. First, we will re-analyze an expanded PSP dataset to estimate fall rates of standing dead trees using Stone et al.ís (2002) approach and including recent measurement data. Second, we will use a retrospective approach to determine species mortality rates over a range of site conditions. This approach will be able to determine which trees are dying, for example, old Sitka spruce or large western redcedar or diseased hemlocks. Retrospective studies are also amenable to static life table analysis, which provides an estimate of fall rates for trees older than those available from the PSP data.

To determine mortality and transition rates of coastal species, we will investigate the general applicability of logistic regression and test other models. We will reconcile the values produced by the two different approaches, PSP and retrospective, and provide updated functions from a broad data source for use in the development of new functions for models such as TASS and its user interface, TIPSY.

By using this combination of approaches, we will be able to examine snag production and dynamics across the full range of stand ages. Such a comprehensive approach is essential to the management of forests to maintain snag habitat for wildlife or other purposes and is important to evaluating losses when estimating stand yields.
Related projects:  FSP_Y071017FSP_Y082017


Executive Summary (73Kb)

To view PDF documents you need Adobe Acrobat Reader, available free from the Adobe Web Site.

Updated August 16, 2010 

Search for other  FIA reports or other Ministry of Forests and Range publications.

Please direct questions or comments regarding publications to