Site-specific nutrient data reveal the importance of soils in driving the Mycorrhizal make-up of woody vegetation worldwide
Résumé
Aim : Arbuscular mycorrhizas (AM) and ectomycorrhizas (ECM) have different impacts on nutrient cycling, carbon storage, community dynamics and enhancement of photosynthesis by rising CO2. Recent global analyses have concluded that patterns of AM/ECM dominance in forests worldwide are shaped by climate, with soil nutrients contributing negligible additional explanatory power. However, their reliance on nutrient data from GIS surfaces masks important local influences of parent material, topography and soil age on soil nutrient status. We asked if use of site-specific nutrient data reveals a more important role for nutrients. Time Period : Present day. Location : Global dataset comprising 703 sites, encompassing forests, savanna/woodlands, shrublands and deserts on all continents except Antarctica. Taxa Studied : Arborescent plants, including angiosperms, gymnosperms and tree ferns. Methods : Generalised Additive Models for Location, Scale and Shape (GAMLSS) to determine the effects of climate variables, soil nitrogen and soil phosphorus on the proportional representation of ECM and of non-mycorrhizal species (NM) in woody vegetation. Results : GAMLSS showed a strong negative relationship of ECM representation with mean annual temperature (MAT), and a strong negative relationship with soil total nitrogen. NM representation was highest on dry sites and phosphorus-poor sites. Reanalysis showed that GIS-derived soil nutrient data had less explanatory power than site-specific nutrient data, and resulted in poorer model fits. Conclusions : Our results support the long-held belief that soil nutrients as well as climate influence the relative fitness of different mycorrhizal syndromes worldwide, and demonstrate the value of using site-specific nutrient data. Soil nutrients should be considered when predicting the impact of climate change on the mycorrhizal composition of vegetation and resulting shifts in ecosystem processes.