Midstory hardwoods are traditionally removed to restore longleaf pine on fire‐excluded savannas. However, recent evidence demonstrating midstory hardwood facilitation on longleaf pine seedling survival has brought this practice into question on xeric sites. Also, midstory hardwoods could facilitate longleaf pine seedling establishment, as hardwood litter may conceal seeds from seed predators or improve micro‐environmental conditions for seedling establishment. However, little is known about these potential mechanisms. In this study, we tracked longleaf pine seed depredation and germination in artificially seeded plots (11 seeds/m2) in a factorial design fully crossing hardwood retention or removal with vertebrate seed predator access or exclusion in the Sandhills Ecoregion of North Carolina, U.S.A. Seed depredation averaged 78% across treatments and was greatest in unexcluded plots. Hardwood retention did not affect seed depredation. Longleaf pine averaged 3.6 germinants/4 m2 across treatments, and was six times more abundant where vertebrates had been excluded. Hardwood removal had a strong positive effect on seedling germination, likely due to the removal of litter, but only when vertebrates were excluded. Our results indicated midstory hardwoods are not facilitating longleaf pine seedling establishment. Nevertheless, our results indicated that hardwood removal may not increase longleaf pine seedling establishment, as seed depredation diminished the effectiveness of hardwood removal under mast seed availability. Collectively, these results demonstrate the underlying complexity of the longleaf pine ecosystem, and suggest that planting may need to be part of the restoration strategy on sites where seed depredation limits longleaf pine natural regeneration.

Questions: While much is known about the impact of tree encroachment on flammability in degraded pine woodlands, little is known about how encroachment is impacting other important ecosystem functions. We investigated how the availability of seed from four encroaching tree species and the presence of a midstorey and litter layer affect seed predator selection. Additionally, we investigated how seed predators, the midstorey, overstorey basal area, substrate availability, and vegetation cover affect germination for a foundational species (Pinus palustris) compared to an encroaching species (Pinus taeda).

This study was established to evaluate underplanting as a method of reestablishing a shortleaf pine (Pinus echinata Mill.) component to a dry upland hardwood stand in the Piedmont region of the southeastern United States. Replicated treatment plots were harvested to retain four levels (approximately 0, 3, 7, and 10 m2 of basal area per hectare) of residual overstory density. One-year-old containerized seedlings with both smaller (93.4 cm3) and larger (113.1 cm3) plugs and bareroot seedlings were underplanted beneath the residual overstory treatments. After five growing seasons, seedling survival averaged 61% and was not meaningfully affected by residual overstory density. Seedling height growth ranged from 1.42 m to 2.61 m and was inversely related to residual overstory density. Containerized seedlings with larger plugs had the highest survival (77.4%) and best height growth (2.11 m), followed by containerized seedlings with smaller plugs (64.3%, 1.76 m) and bareroot seedlings (40.2%, 1.85 m). The results of this study indicated that underplanting containerized seedlings, particularly those with higher plug volume and greater plug depth, was a suitable option for reestablishing shortleaf pine on drier, hardwood dominated upland sites in the Piedmont. However, even low levels of overstory retention suppressed seedling height growth after a few years. Study Implications The study was conducted on a dry upland site typical of the North Carolina Piedmont. Retaining up to 10 m2 ha–1 of oak and hickory overstory basal area did not strongly affect survival among underplanted shortleaf pine seedlings after five growing seasons. However, overstory cover as low as 3 m2 ha–1 had negative effects on height growth of underplanted seedlings over the same time period. Height growth declined as overstory density increased. Containerized seedlings had better survival than bareroot seedlings. Further improvements in survival and height growth were realized by planting containerized seedlings with higher plug volume and greater plug depth.

IntroductionField experiments of gap-based harvest systems in temperate northern hardwood forests have provided inconsistent support for the theory that such regeneration approaches can improve regeneration success among increasingly underrepresented tree species intermediate in shade tolerance. We established a field experiment in Wisconsin, USA testing the long-term response of tree regeneration to group selection harvesting that created small (50 m2–D:H 0.32), medium (200 m2—D:H 0.64) and large (380 m2—D:H 0.88) gaps.Materials and methodsLight levels were modeled to test the assumption of resource heterogeneity within and across these gap sizes. Spatial statistics were used to assess whether the point patterns of saplings of several midtolerant tree species (Betula alleghaniensis, Carya cordiformis, Fraxinus spp.) mapped in harvest gaps provided evidence for partitioning across resource gradients 9 years post-harvest. Finally, we measured occurrence and height of all saplings within the harvest gaps to test the effects of gap size and deer exclusion fencing on the density and total height of shade-tolerant and midtolerant tree regeneration.Results and discussionLight levels progressively increased with gap size, varied by position within gaps, and provided significant evidence of resource heterogeneity. The complex light gradients formed both within and across the three gap sizes may partially explain the statistically significant clustering of Fraxinus spp. saplings in gap centers and north and east into the forested edge of medium and large gaps, C. cordiformis on the west sides of forested transitions surrounding large gaps, and B. alleghaniensis in the centers of large gaps. Densities of tolerant and midtolerant saplings were similar in medium and large gaps after nine growing seasons, though tolerant saplings averaged two meters taller across all three gap sizes. Our results show that gap-based silvicultural systems can create resource gradients that are sufficient to regenerate mixtures of species with high to intermediate tolerance of shade. However, prescriptions beyond harvesting and deer exclusion fencing may be necessary for the recruitment of midtolerant species to canopy positions.

No abstract