| Popis: |
Gap-based silviculture and legacy-tree retention are strategies that can increase structural and resource heterogeneity while maintaining forest productivity at the stand scale. Nevertheless, areas of delayed regeneration have been noted in openings with some persisting for a decade or more. To better understand the causes of localized recruitment delay and/or near-term failure, we examined spatial patterning of persistent (≥15 years) patches of regeneration failure (≤1 sapling ≥1.37 m tall per 10 m2) at two long-term experimental gap experiments in northern Wisconsin and the Upper Peninsula of Michigan. During the summers of 2018 and 2019, we surveyed 84 harvest-created canopy gaps for areas of failure and mapped 79 individual patches where regeneration had failed to establish and recruit in a total of 52 openings. Spatial attributes of these individual patches were summarized and compared to their location relative to expected within (e.g., north versus south, distance from edge) and between gap (e.g., gap size) gradients in light resource availability. Spatial analyses revealed no significant trends between north and south regions of gaps, which suggests that regeneration failure at the gap scale is likely determined by factors disentangled from expected within-gap light resource gradients. At the site without legacy-tree retention, patches that had failed to regenerate trees occupied 17.3% of the total area in gaps (≥10 m radius) and were located near gap center. In contrast, at the site with legacy trees, only 4.8% of the total area in gaps (≥11 m radius) was occupied by patches of regeneration failure and patches were located near gap edge. Furthermore, increasing gap size was associated with a higher risk of regeneration failure only at the study site without legacy-tree retention. Our results suggest that legacy-tree retention may moderate the spatial extent of regeneration failure, however, further work is needed to clarify their long-term influence on regeneration dynamics within harvest openings. |
Full Text from ScienceDirect