Zeta diversity differentiates factors driving community assembly of rare and common ectomycorrhizal fungi

Abstract

A fundamental goal in community ecology is to understand what factors drive community assembly processes. The factors affecting ectomycorrhizal fungal communities are unknown in many regions, particularly in the southern hemisphere. We investigate community assembly using ITS2 metabarcoding of ectomycorrhizal fungi sampled from 3943 hyphal ingrowth bags buried in 81 Nothofagus forests across New Zealand's South Island. By applying zeta diversity analysis and multisite generalized dissimilarity modelling (MS-GDM) we quantify the effects of 43 biotic and environmental variables on community turnover. Unlike traditional beta diversity analyses that are heavily influenced by rare species, the zeta diversity framework differentiates between factors driving turnover of rare and common species, providing a more complete picture of community dynamics. We found that community assembly was dominated by deterministic rather than stochastic processes and identified ecological factors affecting all taxa, as well as others that were specifically important to rare or common taxa. Soil variables were important drivers of turnover for all species, whereas ground cover variables, forest patch size, precipitation and host tree identity had greater effects on rarer species, and tree size and temperature effects were specific to more common species. Interestingly, the effect of temperature on common species is in line with recent evidence from other Kingdoms, pointing to possible generality, and highlighting the importance of considering common species. Applying these methods to fungi has allowed us to identify the distinct ecological processes that structure rare and common taxa during community assembly. This has important implications for understanding the functional effects of community responses to environmental change.

Keywords: Nothofagus (southern beech; Nothofagaceae); community assembly; ectomycorrhizal fungi; environmental filtering; host specificity; multisite generalized dissimilarity modelling (MS-GDM).