Lodgepole pine (Pinus contorta) ecotones in Rocky Mountain National Park, Colorado, USA

Because ecotones often represent the physiological or competitive limit-of-distribution of species, they serve to define a species' local distribution. We take a relatively new approach to gradient analysis to quantify the changes in forest dominance (basal area of dominant tree species) and environmental factors (elevation, slope, aspect, intercepted photosynthetically active radiation [PAR], summer soil moisture, and soil depth and texture) across current lodgepole pine (Pinus contorta) ecotones in Rocky Mountain National Park, Colorado, USA. Based on canonical correspondence analysis on data from eight 20 m wide, 140-480 m long vegetation transects, we found that lodgepole pine distribution, is described generally by elevation and moisture gradients, and to a lesser degree by soil characteristics. The first two ordination axes explained 72% of the variance in forest patterns, Environmental factors significantly correlated to the first ordination axis included: elevation, summer soil moisture, and percentage sand in the soil. Environmental factors significantly correlated to the second ordination axis included: growing-season soil moisture, surface rockiness, percentage silt, and percentage clay. However, superimposed on these general environmental gradients are ecotone-specific gradients that cannot be detected by typical gradient analysis approaches using only homogeneous plots. The basal area of lodgepole pine decreased nonlinearly and more sharply toward the ecotones than did the basal area of the neighboring tree species. Stepwise multiple linear regression and path coefficient analysis models showed that the basal area gradients of lodgepole pine into spruce/fir (Picea engelmannii, Abies lasiocarpa) forests correlated with summer soil moisture; lodgepole pine gradients into lower elevation ponderosa pine (Finns ponderosa) forests correlated strongly with intercepted PAR, slope, percentage silt, soil depth, and percentage clay; and lodgepole pine gradients into limber pine (Pinus flexilis) forests correlated largely with elevation, intercepted PAR, percentage sand, and percentage clay. Furthermore, the factors controlling tree species distributions at ecotones are not symmetrical. For example, at the lodgepole pine to limber pine ecotone, different factors or different magnitudes of the same factors determined the upper limit of lodgepole pine and the lower limit of limber pine. We conclude that: (1) different factors control a species' upper and lower elevation limits: (2) unequal competition for resources occurs between tree species where their ranges overlap; and (3) generally, soil differences may not substantially restrain the movement of some forest types into neighboring types.