Riparian vegetation response to gradients in residual basal area with harvesting treatment and distance to stream

The long-term sustainability of riparian management zones (RMZs) depends upon the maintenance of desirable ecosystem conditions within the riparian buffer. We assessed the change in understory vegetation (density and/or biomass of tree seedlings, saplings, woody shrubs, and herbs) in response to different levels of residual basal area within eight northern hardwood RMZs in northeastern Minnesota, USA. Three-year herbaceous cover, diversity, and species composition was investigated in a subset of four of these sites. RMZs were 46m wide and located next to upland clearcuts. Three overstory harvest treatments were applied in the RMZs: uncut controls with an average overstory basal area of 23m ²ha ^-1, overstory basal area reduction to 16m ²ha ^-1 in the medium-retention RMZs, and overstory basal area reduction to 8m ²ha ^-1 in the low-retention RMZs. Both partial cut treatments followed best management practices to decrease management intensity towards stream banks, resulting in a gradient of decreasing residual basal area corresponding to increasing light availability with increasing distance from the stream. Understory shrub and sapling biomass, and density of small woody regeneration, including aspen/birch and beaked hazel, all increased in the partially harvested treatments by the third year following treatment, particularly in the low-retention RMZs. In all treatments, along the gradient of decreasing basal area with distance from the stream, small woody biomass and aspen/birch regeneration densities responded positively with distance to stream and increased light availability. Species with higher light requirements increased significantly with more light availability. In contrast, herbaceous species with higher nutrient and heat requirements responded negatively to increased light availability and increased distance to the stream, including in the controls. This response was large due to species differences between the location nearest to the stream and those further from the stream and most likely reflects different species compositions among fluvial landforms. While both partial harvesting treatments resulted in aspen regeneration densities within acceptable ranges for timber production standards, the lower (though not statistically significant) beaked hazel densities following the more conservative partial harvest treatment might be more favorable for the development of longer-lived conifer species in the long run. Further, many of the herbaceous responses were only observed at the extreme ends of the treatment gradient. Regardless of treatment, the edge effect on light availability, which extended at least 5m into the uncut control RMZ, may reduce the functional width (i.e., effectiveness) of the vegetation buffer.