Understory plant community classifications as predictors of forest site quality for lodgepole pine and white spruce in west-central Alberta
Understory vegetation of 103 lodgepole pine (Pinuscontorta Loudon var. latifolia Engelm.) and white spruce (Piceaglauca (Moench) Voss) stands, 70–150 years old, in a 16 000 km2 region of west-central Alberta was classified by 14 methods using species cover or log2 cover classes. Six classifications, including five selected by dendrogram and ordination elucidation of understory community types (UCTs), were evaluated by analysis of variance to identify UCTs with significantly different site indices of pine and spruce. Each classification consists of three to five UCTs, referable to two forest types, both levels distinguished by differences in constancy and cover of understory species. Significant differences (P < 0.05) in pine site index at 70 years occur between UCTs belonging to the same and different forest types, using divisive two-way indicator species analysis (TSL) and agglomerative furthest neighbour (FNL) and minimum variance (MVL) cluster analysis classifications, all based on log2 cover classes. Comparable differences between UCTs for spruce occur within one forest type, using FNL, MVL, and minimum variance cluster analysis based on raw percent cover (MVR). Ordination of stands based on understory species cover reveals that spatially segregated UCTs usually have significantly different site indices. If a single classification method is desired for both lodgepole pine and white spruce, MVL appears most suitable for predicting site indices in the region.