
| HOME SCHEDULE AUTHOR INDEX SUBJECT INDEX |
|
130 A cross-site comparative study of physical correlates with community transition. LOOKINGBILL, TODD1, PIERCE, KENNETH1, URBAN, DEAN1, 1 ABSTRACT- Assessing vegetation pattern across environmental gradients is a fundamental challenge to plant ecology. Analytic techniques in ordination and classification can be used in synthesis to identify broad scale patterns in forest communities. We applied a common field sampling protocol to four forested, montane sites with contrasting climates, terrain and geologic history: Sequoia National Park (Sierra Nevada, California), H.J. Andrews Experimental Forest (Cascades, Oregon), Hubbard Brook Experimental Forest (White Mountains, New Hampshire), and Coweeta Hydrologic Laboratory (Southern Appalachians, North Carolina). A series of nonmetric multidimensional scaling (NMS) and clustering analyses were conducted to evaluate general relationships between species composition and the environmental template at the four sites. Elevation was identified as the primary correlate with species pattern at Sequoia, Andrews, and Hubbard Brook. The strength of correlation varied with the steepness of the gradient (Sequioa r2 = 0.73 with the strongest NMS axis; Andrews r2 = 0.56; Hubbard Brook r2= 0.37). Additional topographic variables were associated with community transitions at high elevation in the western sites. Soil properties appear to be more important controls at Coweeta. Soil texture, base saturation, and pH were major correlates, though none of the variables were as tightly coupled with the species abundance data as at the western sites. The value of cross-site comparative studies for identifying generalities and site-specific responses to environmental gradients is discussed. KEY WORDS: landscape scale, multivariate analysis, environmental gradients |