The initial species composition and conditions following disturbance are crucial to determining the long-term development of vegetation (Egler 1954, Glenn-Lewin et al. 1992). Thus, an understanding of early succession can provide insights into the mosaic of vegetation types that will emerge on a landscape. Observations in the Rocky Mountains and elsewhere indicate that postfire vegetation patterns vary considerably. It is now known that numerous factors simultaneously contribute to the species composition of pioneer vegetation (Stahelin 1943, Cattelino et al. 1979, Glenn-Lewin 1980, Noble and Slatyer 1980, Keever 1983, Walker and Chapin 1987, Pickett 1989, Westoby et al. 1989, Glenn-Lewin et al. 1992), and often different patterns of early succession are possible even on sites that are similar. While research on the distribution of late successional vegetation along environmental and disturbance gradients has been a dominant theme in plant ecology (Daubenmire 1943, Whittaker 1956, Despain 1973, Loope and Grueii 1973, Whipple and Dix 1979, Peet 1981, Romme and Knight 1981, Wentworth 1981, Veblen 1986, Allen and Peet 1990), few studies have directly investigated the factors that affect the distribution of pioneer plant communities. Moreover, the influence of landscape structure on the development of vegetation mosaics has only recently been considered (Milne and Forman 1986, Turner et al. unpublished manuscript). Studies designed to understand vegetation patterns following disturbance could be improved by considering site history and the character of the surrounding landscape as well as the effects of site features. We have initiated research to consider the effect of various spatial, historical and site factors on early postfire vegetation in Grand Teton National Park (GTNP) and the adjcent Bridger-Teton National Forest. Specifically, we are studying the relationships between characteristics of early succession and fire severity, geologic substrate, soil characteristics, mean burn patch size, prefire canopy composition, cone serotiny, topographic position, distance to unburned vegetation, and the vegetation composition and fire severity of the surrounding landscape. Our research objectives are to: 1) determine the range of variability in the patterns of early postfire succession 2) determine which environmental and historical variables are most important in controlling vegetation development following fire 3) elucidate the importance of landscape position and between-patch interactions during early postfire succession 4) develop a series of multivariate models that will predict the characteristics of early succession in different situations; 5) project, based on existing literature, the most probable trajectories of each pioneer community type. An important component of our research is to examine the influence of landscape context on early successional vegetation. Understanding and predicting the behavior of organisms requires the consideration of a broad environmental matrix, not simply the autecology of the organisms (Rowe 1981). According to Glenn-Lewin and van der Maarel (1992), spatial attributes such as landscape complexity, community isolation, and the spatial scale of disturbance are important factors influencing succession. We will use GRASS-GIS, to better understand the relationship between site characteristics and pioneer vegetation pattern (Davis and Dozier 1990, Davis and Goetz 1990) as well as to elucidate the importance of landscape position in early succession.
Read full abstract