Rocky Mountain Research Station, USDA Forest Service, Fort Collins, Colorado, USA

Background and objectives
Most of what we know about the impact of forest diseases comes from studies of catastrophic conditions. Studies of outbreaks often deal with the direct relationship between disease and tree mortality. This one to one direct relationship is less clear during non-outbreak conditions, primarily because pathogens interact with so many other elements of the ecosystem. It is difficult to determine what single agents do because it is difficult to differentiate their direct impact from the indirect impacts associated with their interactions with other agents. We usually either make no distinction between direct and indirect effects, or we focus only on the direct effects. This has been a major problem in developing useful forest disease assessment systems or predictive models. The aim of this study was to examine the influence on forest stand development of Armillaria root disease relative to other small-scale disturbances. The specific objectives were: (i) to characterize and quantify the relationship of diseases and associated mortality agents on stand development, and (ii) to develop a hypothetical model of stand development linked to disease interactions.

Materials and methods
Transect surveys were used to map out the stand distribution within the study area [1]. Each stand was distinguished by cohort type. For this study, cohort was defined as, "an aggregate of trees that starts as a result of a single disturbance". Each cohort was assessed visually for tree killing agents. A stand development model was hypothesized that suggested the relationship between cohort structure and various disturbances. The direct and indirect effects of Armillaria root disease were calculated using path analysis. Path analysis is a method of assessing covariation among ecosystem elements and can be used to estimate the relative strengths of direct and indirect interactions among these elements.

Results and conclusions
Stand structure posed constraints on the activity of Armillaria; thus, this disease acted only in certain windows of time and space. The nature, magnitude, and intensity of Armillaria and other disturbances were reflected in the structure and development of the study area, which resulted in a patch mosaic of tree cohorts that varied in canopy structure, stem size, stem density, and age. Each cohort represented a different stage of stand development and was associated with a specific suite of disturbances. The stand initiation stage (0-71 yrs) was dominated by western gall rust, ice and snow damage, self-competition, and self-suppression. The stem exclusion stage (72-171 yrs) was associated mainly with mountain pine beetle. The understory reinitiation stage (172 - 200 yrs) had either low intensity wildfires, shrub competition, low light intensity and limb rust, or mountain pine beetles, windstorms, and lightning, or Armillaria root disease. Stand replacement (200 - 300 yrs) resulted from mountain pine beetle, wildfire, or windstorms. The ecological role of a tree killing agent depended on its relative position in time and space in the hierarchical structure of the forest ecosystem. Disturbances play basically two types of roles: 1)regenerators of new stands, or 2) catalysts of stand development. Armillaria root disease acted mostly as a catalyst of stand development, had the largest direct effect where it acted alone, interacted mostly with wind, and had a negative effect on beetle activity.

1. Lundquist JE, 1995. Forest Ecology and Management 74, 37-48.