Mycorrhizal fungi play key roles mediating carbon and nutrient cycling in forest ecosystems. In eastern North America, temperate forests have mixed assemblages of arbuscular and ectomycorrhizal trees species, but these types differ in the physiology, biochemistry, and traits of both host and fungal species. Over long time scales, such differences may lead to divergent patterns in soil biogeochemistry, and soil stocks of carbon and nitrogen in particular. Relative to arbuscular mycorrhizal trees, ectomycorrhizal trees are hypothesized to increase the amount of carbon stored per unit nitrogen in soils. However, it is unclear whether these mycorrhizal effects override those of climate and plant phylogeny. In this study we analyzed Forest Inventory and Analysis data covering temperate forests in the eastern half of the United States. We assessed patterns of dominance by arbuscular and ectomycorrhizal tree hosts across and used a spatially explicit model to relate their distribution to percent content of soil carbon and soil nitrogen in multiple soil horizons (organic, 0-10 cm, and 10-20 cm). We further considered the identity of the tree hosts as being either angiosperm or gymnosperm and determined their effects on patterns of soil carbon and nitrogen.
Arbuscular and ectomycorrhizal trees were well distributed throughout the study area, but ectomycorrhizal trees were dominant in forests at lower and higher latitudes while arbuscular mycorrhizal trees were dominant in those of mid latitudes. Soil carbon to nitrogen ratios increased with increasing dominance of ectomycorrhizal trees, however, this pattern varied by soil horizon. Relative to forests dominated by arbuscular mycorrhizal trees, those dominated by ectomycorrhizal species were associated with greater soil carbon in the upper mineral soil layer (0-10 cm), but not in the lower mineral soil layer (10-20 cm). Moreover, the ectomycorrhizal pattern of greater soil carbon to nitrogen ratios in the organic horizon a primarily explained by higher abundance of gymnosperms over angiosperms. At broader scales (i.e. 10x10 km grid cells), and using a combined soil profile to 1 m depth, we no longer detected differences in soil carbon and nitrogen between mycorrhizal types. Our findings suggest a predictable, localized effects of mycorrhizal association on soil carbon and nitrogen across forests of the eastern United States. However, our results indicate that these relationships can vary within soil profiles and be strongly affected by host plant phylogeny.