Document Type

Yellowstone National Park Report

First Page


Last Page



The importance of large, infrequent natural disturbances, particularly as they influence post­disturbance succession, is well recognized. The conceptualization of ecological systems and landscapes as mosaics of patches generated by disturbance emphasizes successional change, which in turn is a major component of ecologists' understanding of disturbance in a landscape context. However, paradigms of succession largely evolved for small-scale dynamics. Subsequently, these paradigms do not adequately predict the spatial variability that occurs within the perimeter of large, infrequent disturbances, nor do they adequately consider the effects of heterogeneity on recovery of the system. Understanding the nature of the disturbance mosaic and the factors controlling landscape patterns are crucial for predicting ecosystem dynamics and vegetation development in disturbance-prone landscapes, and research that addresses these questions remains a priority. Landscape heterogeneity following large, mfrequent fires is represented by a patchwork of burned and unburned vegetation as well as by a mosaic of burn severities (Turner et al. 1997, Romme et al. 1998), so that the mosaic of stand structure and function produced by fire is a legacy of the disturbance itself as well as that of the pre-fire forest. In this sense, large, infrequent fires are thought to impose a persistent influence on landscape pattern. Landscape pattern has large implications for biodiversity, determines the connectivity of habitat, affects the spread of disturbances such as insect outbreaks, and may influence the initiation and spread of small, frequent surface fires by affecting the spatial distribution of fuels. Similarly, spatial heterogeneity in stand structure that exists across a landscape may influence the propagation of crown fires (Van Wagner 1977, Turner and Romme 1994). Clearly, understanding the changes that occur in landscape pattern provides the key to understanding the dynamics of many ecological processes. Our study examines the natural changes that occur in landscape pattern with succession in Yellowstone National Park (YNP) between large, infrequent wildfires. The 1988 fires in YNP created a mosaic of burn severities that produced tremendous variation in density of lodgepole pine (Pinus contorta var. latifolia) seedlings across the landscape (Turner et al. 1994). Because the Yellowstone mosaic of post-fire seedling densities varies from high-density stands (>50,000 trees/ha), which will inevitably be subject to self-thinning, to low-density stands (<500 stems/ha) which have the potential to be colonized by seed sources from outside the post-fire stand (Turner et al. 1997), tree density in initially dissimilar stands may potentially converge with time, thereby affecting landscape pattern. Although investigators have speculated that the post-fire mosaic of initial successional pathways will persist until the next large, infrequent fire (Anderson and Romme 1991, Ellis et al. 1994, Wallin et al. 1994, Turner et al.1997), little data exists to quantify the degree of persistence of this landscape legacy. In the summer of 2000, we initiated a study of the long-term development of the individual patches (forest stands) that constitute the mosaic in order to understand the degree to which a major disturbance will determine landscape pattern over the long term. The data collected during that period continues to be analyzed and incorporated into modeling efforts and will be supplemented with additional field data collected during the summer of 2001. Our study addresses two major questions in the context of a disturbance-driven forest landscape: i) Do the initial differences in stand density that exist immediately after stand-replacing fires converge with stand development, and what are the rates and mechanisms of this convergence?; and ii) Based upon empirical data of stand dynamics for YNP, how is the spatial arrangement of stands of differing structural characteristics likely to change before the next large, infrequent disturbance?