No abstract

Forested wetlands are an integral but understudied part of heterogeneous landscapes in Atlantic Canada, although they are known to provide habitat for species at risk. Our objectives were to explore patterns of forest structure across edges between forested wetland and upland forest, to locate changes in vegetation structure and to assess multivariate relationships in vegetation structure. Our study sites were in temperate (Acadian) forested wetland landscapes. We sampled trees and recorded canopy cover every 20 m along 120 m long transects. We estimated the cover of trees, saplings, shrubs in three height classes, Sphagnum, other bryophytes, lichens, graminoids, ferns and forbs in contiguous 1 x 1 m quadrats. We calculated structural diversity using the Shannon index and used wavelet analysis to assess spatial patterns. We found few clear patterns except for lower tree structural diversity at the edge of forested wetlands. Structural diversity was not a reliable measure for distinguishing forested wetland from upland forest. Forested wetlands are an integral part of many forested landscapes in Atlantic Canada but their detection and differentiation from surrounding ecosystem can be difficult. Policy should err on the side of caution when mapping forested wetlands and include them in wetland protection.

The lichens Erioderma pedicellatum (Hue) P.M. Jørg. and Erioderma mollissimum (G. Sampaio) Du Rietz are endangered species in Canada. Both species are obligate epiphytes found in forested wetlands near the Atlantic Coast. They are thought to be primarily threatened by logging and acid pollution, but the influence of these factors has not been examined at large-extents or relative to other habitat features. Critical habitat for protection has remained difficult to define beyond observed occurrences, because of the low accuracy of existing habitat models. To facilitate improved recovery planning and understanding of their ecology in the province of Nova Scotia, we created high-resolution distribution models for both species, incorporating elements of climate, forest composition, hydrology, acid pollution, and anthropogenic influence using the MaxEnt algorithm and a backwards stepwise selection process. The most important predictors were related to rainfall or an oceanic moderation of thermal optima. Depth to water table and the presence of suitable forest composition were also included, as was distance from roads for E. pedicellatum. The putative threats, acid pollution and silvicultural treatment, were not important and therefore excluded from models. Although both species have highly specific habitat requirements, E. pedicellatum appears to be more sensitive to human activities.

No abstract