View article

In studies of edge effects, it is difficult to separate edge from size effects because size and edge are strongly correlated in most cases. In our study we separated these two effects by examining small forest patches of two different size classes ( <1 ha and 4–6 ha ), with patches chosen so that area and perimeter were not correlated within each class. We conducted our study in a mosaic of old‐growth Picea abies forest and wetland, consisting of forested moraine hills ( “islands” ) in a Sphagnum‐bog matrix. On each island, we demarcated a 0.1‐ha sample plot and measured site and forest characteristics. In each plot, we examined all fallen logs for the occurrence of epixylic hepatics ( liverworts ), and all standing trees for the occurrence of calicioid lichens. We examined correlations between species and area, shape, and distance to nearest edge. The effect of shape on forest‐interior conditions was analyzed …