Plant-soil feedbacks are known to play a central role in species co-existence, but conceptual frameworks for predicting their magnitude and direction are lacking. We ask whether co-occurring trees that associate with different types of mycorrhizal fungi, which are hypothesized to differ in terms of nutrient use and plant-soil feedbacks, differ in sapling establishment densities and probability of co-occurrence. Given that ectomycorrhizal (ECM) trees typically have fungal structures that protect roots from pathogens whereas arbuscular mycorrhizal (AM) trees do not, we hypothesized that ECM saplings would be clustered around ECM trees, while AM saplings would be suppressed near AM trees. Most previous studies have focused on seedlings, but here we examine whether the spatial signal is evident in later life stages. We measured the spatial associations of ~ 28,000 trees using point pattern analysis in a 25-ha old-growth forest where ECM trees comprised 72% of total basal area and 42% of the total stems, while AM trees comprised the remainder. Supporting our hypothesis, AM saplings were more inhibited by AM trees, while ECM saplings were more clustered around ECM trees. The spatial patterns of AM and ECM trees on saplings of the alternate mycorrhizal type were inhibited. To the extent that similar types of feedbacks occur for other AM and ECM trees, our results suggest that fundamental differences in the nature of local-scale biotic interactions between trees and their fungal symbionts may influence forest community assembly and ecosystem dynamics.
Keywords: Arbuscular mycorrhizae; Ectomycorrhizae; ForestGEO; Oak–hickory forest; Pair-correlation function.