Arbuscular mycorrhizal (AM) fungi are integral components of grasslands because most plants are associated with interconnected networks of AM hyphae. Mycorrhizae generally facilitate plant uptake of nutrients from the soil. However, mycorrhizal associations are known to vary in their mutualistic function, and there is currently no metric that links AM functioning with fungal colonization of roots. Mycorrhizal structures differ in their physiological and ecological functioning, so changes in AM allocation to intraradical (inside roots) and extraradical (in soil) structures may signal shifts in mycorrhizal function. We hypothesize that the functional equilibrium model applies to AM fungi and that fertilization should reduce allocation to arbuscules, coils, and extraradical hyphae, the fungal structures that are directly involved in nutrient acquisition and transfer to plants. This study compared AM responses to experimental N enrichment at five grasslands distributed across North America. Samples were collected from replicated N‐enriched (and some P‐enriched) and control plots throughout the growing season for three years. Intraradical AM structures were measured in over 1400 root samples, extraradical hyphal density was measured in over 590 soil samples, and spore biovolume was analyzed in over 400 soil samples. There were significant site × N interactions for spore biovolume, extraradical hyphae, intraradical hyphae, and vesicles. Nitrogen enrichment strongly decreased AM structures at Cedar Creek, the site with the lowest soil N:P, and it increased AM structures at Konza Prairie, the site with the highest soil N:P. As predicted by the functional equilibrium model, in soils with sufficient P, relative allocation to arbuscules, coils, and extraradical hyphae was generally reduced by N enrichment. Allocation to spores and hyphae was most sensitive to fertilization. At the mesic sites, this response was associated with a shift in the relative abundance of Gigasporaceae within AM fungal communities. This study demonstrates that N enrichment impacts mycorrhizal allocation across a wide range of grassland ecosystems. Such changes are important because they suggest an alteration in mycorrhizal functioning that, in turn, may impact plant community composition and ecosystem function.

Arbuscular mycorrhizal fungi (AMF) are considered both ecologically and physiologically important to many plant communities. As a result, any alteration in AMF community structure following soil nitrogen (N) enrichment may impact plant community function and contribute to widespread changes in grassland productivity. We evaluated the responses of AMF communities to N fertilization (≥100 kg N·ha−1·yr−1) in five perennial grasslands within the Long‐Term Ecological Research network to generate a broader understanding of the drivers contributing to AMF species richness and diversity with increasing soil N fertility, and subsequent effects to host‐plant communities. AMF spore and hyphal community data at three mesic sites (Cedar Creek, Kellogg Biological Station, Konza Prairie) and two semiarid sites (Sevilleta, Shortgrass Steppe) were collected over two consecutive years and used to test four hypotheses about AMF responses to N fertilization. Under ambient soil N, plant annual net primary productivity and soil phosphorus (P) were strongly related to climatic differences in AMF communities (semiarid vs. mesic). Following N fertilization, the drivers of AMF community structure were soil N availability, N:P supply ratio, and host‐plant photosynthetic strategy (C3 vs. C4) but not climate. In P‐rich soils (low N:P), N fertilization reduced AMF productivity, species richness, and diversity and intensified AMF community convergence due to the loss of rare AMF species and the increased abundance of Glomus species. In P‐limited soils (high N:P), AMF productivity, species richness, and diversity increased with N fertilization; the most responsive AMF taxa were Acaulospora, Scutellospora, and Gigaspora. Soil N or N:P × host‐plant (C3, C4) interactions further modified these responses: AMF hyphae (primarily Gigasporaceae) associated with C3 plants increased in abundance with N fertilization, whereas C4 plants hosted nitrophilous Glomus species. Such responses were independent of the duration or quantity of N fertilization, or the time since cessation of N fertilization. This synthesis provides a new understanding of AMF community patterns and processes, and it identifies three key drivers (soil N, N:P, host plant) of AMF community structure that may be tested in other communities.