Ecological classification of wetlands in the western half of the Hiawatha National Forest, Upper Michigan, United States

Wetlands comprise newly half of the total land am in the western half of the Hiawatha National Forest, Michigan. The characteristics of these wetlands am poorly understood and no regional classification existed that could address management needs. Ecosystem classifications that an based on multiple biotic and abiotic factors with an understanding of their interrelationships are recognized as the most useful for multiple use management. The objectives of this study were to (1) develop a multifactor ecological classification system for the Forest, (2) characterize the interrelationships between wetland biotic and abiotic components, (3) investigate peatland vegetation in relation to groundwater characteristics and classification concepts of bog, fen, and mop, (4) mid validate National Wetlands Inventory (NWI) maps. A diverse set of biotic and abiotic variables were measured on 118 plots.;A wetland classification was created with 23 ecological landtype phases (ELTPs). ELTPs are repeatable and consistent landscape waits with similar landform, vegetation, hydrology, and soil. Wetland ELTPs were grouped into two soil landtypes, organic (>30 can organic depth) and mineral. Each landtype was subdivided into four subgroups: two dichotomies of nonforested or forested and acidic or base-rich. The subgroups were based on the two primary hydrogeomorphic gradients identified during the study (1) open-to-forested and (2) acid-to-allailine. Species richness for bryophytes, herbaceous, and shrub/tree physiognomic, groups was also correlated to these gradients but each group differed in their relationship to individual factors. Surface structural variables like woody debris and microtopography were more important for bryophytes than herbs or shrub/trees.;NWI mapping of upland and wetland areas was over 90% accurate when compared with ground truth data. Accuracy was lowest with forested wetlands. The most common NWI error was the classification of wetlands on the upland AuGres soil series.;A DCA ordination with a Spearman rank-correlation analysis related peatland vascular and bryophyte vegetation patterns most strongly to groundwater Ca, Mg, and pH. A discriminant analysis was able to successfully identify bog, fens, or swamps at an overall 80% accuracy level.