Figure 1. Patterns of root-derived C accumulation related to (a) % ECM tree dominance (by basal area) of the plot and (b) root production. Each point represents one plot, with distinct sites depicted with different shaped points. Sites include Harvard Forest (HF), Lilly-Dickey Woods (LDW), Smithsonian Conservation Biology Institute (SCBI), Smithsonian Environmental Research Center (SERC), Tyson Research Center (TRC), and Wabikon Lake Forest (WLF). There is a weak negative relationship between root-derived C accumulation and plot % ECM (R2 = 0.09, P = 0.013) but no relationship between root-derived C accumulation and root production at a given plot. Significant linear regression shown with shaded 95% confidence interval.

Figure 2. (a) Root-derived C accumulation and (b) root production related to plot mycorrhizal type. Bars represent means (with standard error bars shown) for the AM-dominated (red) and ECM-dominated (blue) end-member plots at each site. Sites include Harvard Forest (HF), Lilly-Dickey Woods (LDW), Smithsonian Conservation Biology Institute (SCBI), Smithsonian Environmental Research Center (SERC), Tyson Research Center (TRC), and Wabikon Lake Forest (WLF). Insets show AM vs. ECM cross-site means, with * indicating significant difference between mycorrhizal types for root-derived C accumulation (P = 0.016). Mycorrhizal type, but not site or the interaction between mycorrhizal type and site, significantly predicted root-derived C accumulation (panel a; P = 0.014). Mycorrhizal type and site significantly predicted root production (panel b; P = 0.04 and P < 0.001, respectively), while the interaction between mycorrhizal type and site was marginally significant (P = 0.09 ). There were no significant pairwise differences between mycorrhizal types for any site in either panel a or b.

Figure 3. Ratio of annual root-derived C accumulation to annual aboveground litter mass C at the plot level for the AM (red) and ECM (blue) end-member plots at each site. Sites include Harvard Forest (HF), Lilly-Dickey Woods (LDW), Smithsonian Conservation Biology Institute (SCBI), Smithsonian Environmental Research Center (SERC), Tyson Research Center (TRC), and Wabikon Lake Forest (WLF). The 1:1 line is shown as a dashed grey line for reference. Bars represent plot-level means (with standard error bars shown) for each mycorrhizal type at each site. Site and mycorrhizal type (but not the interaction between the two) significantly predicted annual root-derived C accumulation: litter mass C (P = 0.003).

Figure 4. Mean root-derived MAOM-C (mg C/ g bulk soil) for the three AM-dominated (red) and three ECM-dominated (blue) end-member plots at three sites with standard error bars (n = 3 for each group). Mycorrhizal type (P = 0.04) significantly predicted plot-level root-derived MAOM-C, and the inset shows cross-site differences in AM vs. ECM mean root-derived MAOM-C across all sites. * Indicates significant differences between mycorrhizal types (P < 0.05). Sites include Harvard Forest (HF), Lilly-Dickey Woods (LD), and Smithsonian Environmental Research Center (SERC).