Abstract

Groups of ground-flora species indicating similar habitat conditions, termed ecological species groups, were studied over the rather narrow range of upland environmental conditions of the Cyrus H. McCormick Experimental Forest, north-central Upper Michigan, United States. Nonparametric and multivariate statistical techniques were used to analyze frequency of occurrence data and contrast the relative indicator value of the groups and species. Species distributions were finely adapted to the local environmental situation. It was possible to identify specific soil drainage, texture, and fertility levels using the ecological species groups with a known probability of error. For example, presence of the Osmunda species group indicated a 90% probability of soil mottling within the upper 40 cm. By observing the presence and absence of the Clintonia and Osmunda species groups it was possible to classify soil drainage (moist conditions) at the 5% level of significance. Similarly, presence of the Viola species group indicated significantly higher soil pH and total nitrogen levels. The combined approach of studying species distributions along single- and multi-factor environmental gradients facilitated understanding of the specific indicator value of the individual ecological species groups. Results demonstrate the ability of the groups to speed the process of soil and site mapping. Because of the demonstrated specificity of species and changes in soils, climate, and composition from place to place, it is probably necessary to systematically reaffirm specific plant-environment relationships when using plants as indicators of forest potential.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call