Discerning responses of down wood and understory vegetation abundance to riparian buffer width and thinning treatments: an equivalence–inequivalence approach

Abstract

The combined effectiveness of thinning and riparian buffers for increasing structural complexity while maintaining riparian function in second-growth forests is not well documented. We surveyed down wood and vegetation cover along transects from stream center, through buffers ranging from <5 to 150 m width into thinned stands, patch openings, or unthinned stands of 40- to 65-year-old Douglas-fir ( Pseudotsuga menziesii (Mirb.) Franco) forests in western Oregon, USA. Small-wood cover became more homogeneous among stream reaches within 5 years following thinning, primarily due to decreases for reaches having the greatest pretreatment abundance. Mean shrub cover converged, predominantly because of decreases in patch openings. Herbaceous cover increased, particularly in patch openings. Relative to unthinned stands, herbaceous cover was similar in wide buffers and increased in the narrowest buffers and in narrow buffers adjacent to patch openings. Moss cover tended to increase in thinned areas and decrease in patch openings. Both conventional point-null hypothesis tests and inequivalence tests suggested that wood and vegetation responses within buffers of ≥15 m width were insensitive to the treatments. However, inherently conservative equivalence tests infrequently inferred similarity between thinned stands or buffers and untreated stands. Difficulties defining ecologically important effect size can limit the inferential utility of equivalence–inequivalance testing.