Factors influencing seed germination can strongly influence community assembly in forest ecosystems. Throughout the southeastern United States, decades of fire suppression have enabled the establishment of fire-sensitive tree species within pine savannas, which has increased tree seed diversity at the forest floor. Contemporary restoration efforts attempt to promote longleaf pine (Pinus palustris) (PP) germination by removing competing vegetation and exposing mineral soil, as it is considered a foundational species in savannas. However, little is known about the indirect effects of restoration on its fire-sensitive competitors.
In this fully crossed factorial experiment, we investigated the effects of midstory hardwood litter and cover (removed/not removed) and vertebrate seed predators (excluded/not excluded) on seed depredation of PP, P. Taeda (PT), Liquidambar styraciflua (LS), Acer rubrum (AR) and Quercus nigra (QN) across gradients of overstory basal area and Aristida stricta (AS) density in 72 hour cafeteria trials in the Sandhills Ecoregion of North Carolina, U.S.A. Additionally, we artificially seeded (12 seeds/m2) each treatment with PP and PT and tracked germination over five months.
Results/Conclusions
Overall, seed depredation averaged 13%. Species differed significantly in their rate of depredation (P < 0.0001), with depredation ranging from 25% of seeds for LS to 6% for AR. Depredation of PP (13%) and PT (12%) did not significantly differ. Depredation varied significantly among treatments for PP (P = 0.0158), which experienced higher depredation where midstory hardwoods were removed and vertebrate seed predators were not excluded (21%) compared to any other treatment. In contrast, germination of PP (14 germinants/m2) and PT (9 germinants/m2) was highest where midstory hardwoods were removed and vertebrate seed predators were excluded (P < 0.0001). Accordingly, PP and PT germination was positively associated with percent cover of mineral soil (ρ = 0.3762, P = 0.009) and AS (ρ = 0.3256, P = 0.002) and negatively associated with hardwood (ρ = -0.5116, P = 0.0034) and pine litter (ρ = -0.3673, P = 0.0026) cover. Collectively, our results indicate that seed predators intermediately select PP and PT and that restoration treatments increase seed depredation on PP. Moreover, PP and PT responded favorably to mineral soil and AS cover, indicating that restoration treatments promote the establishment of both species setting the stage for future competition.