We evaluated the total shoot biomass of experimental community pots and found negative feedbacks derived from additions of redcedar roots (Fig. 1). In sterilized field soils, the 400 mL root-addition treatment resulted in a substantial reduction in biomass relative to the control group (Hedges’ g: -2.26; 95% CI: -3.65 to -1.16). In live field soils, shoot biomass declined with each level of root addition, with the 400 mL treatment group having the largest effect (Hedges’ g: -6.53; 95% CI: -7.93 to -5.29). Biomass in potting soil was reduced at the
200 mL (Hedges’ g: -1.38; 95% CI: -2.4 to -0.47) and 400 mL (Hedges’ g: -1.53; 95% CI: -2.57 to -0.66) root additions. For the pairwise comparison of shoot biomass between live and sterile field soils at the 0 mL root treatment levels, cumulative plant biomass in live soils was substantially smaller than in sterile field soils (Figure S1). However, we cannot assert the soil microbial community alone is responsible for observed differences when evaluating raw biomass. Soil testing revealed the sterilization procedure by autoclave released nutrients (notably, phosphorus) that were not available in live field soils (Table S1). For this reason, we did not consider differences between live and sterilized soil treatments derived from raw biomass as being caused solely by the soil microbial community.
We assessed the raw shoot biomass of each species grouped by soil type and root addition treatment and found mixed feedbacks (Fig. 2). In live field soils, B. inermis shoot biomass was reduced in the 200 mL (Hedges’ g: -1.29; 95% CI -2.15 to -0.17) and 400 mL (Hedges’ g: -1.73; 95% CI -2.74 to -0.68) root additions. There was no detected difference in biomass between root treatments in sterile field soils for B. inermis. There was a reduction in biomass of B. inermis at the 400 mL (Hedges’ g: -1.11; 95% CI -1.93 to -0.18) root-treatment level in potting soil. In live field soils, E. canadensis shoot biomass was reduced in the 400 mL (Hedges’ g: -2.58; 95% CI -3.73 to -1.57) root addition. In sterile soil, E. canadensis shoot biomass was reduced in the 200 mL (Hedges’ g: -0.98; 95% CI -1.84 to -0.09) and 400 mL (Hedges’ g: -1.43; 95% CI -2.46 to -0.49) root addition. Shoot biomass of E. canadensis was reduced when grown in potting soil at the 200 mL (Hedges’ g: -0.9; 95% CI -1.61 to -0.07) and 400 mL (Hedges’ g: -1.05; 95% CI -1.72 to -0.26) root additions. We detected positive feedback for B. curtipendula biomass grown in sterile field soil at the 200 mL (Hedges’ g: 1.24; 95% CI: 0.49 to 1.77) and 400 mL (Hedges’ g: 1.76; 95% CI 0.79 to 3.39) root additions. In contrast, B. curtipendula biomass grown in live field soil had a reduction in biomass at the 200 mL (Hedges’ g: -1.20; 95% CI: -2.14 to -0.25) and 400 mL (Hedges’ g: -1.70; 95% CI -2.78 to -0.69) root additions. There was no detected difference in the biomass of B. curtipendula among root treatments in potting soil. Our analysis found positive feedback for S. scoparium grown in sterile field soil with increased biomass at the 100 mL (Hedges’ g: 4.14; 95% CI: 2.67 to 6.46), 200 mL (Hedges’ g: 1.83; 95% CI: 0.79 to 2.50) and 400 mL (Hedges’ g: 1.51; 95% CI: 0.78 to 2.16) levels of root addition. Conversely, the biomass of S. scoparium was reduced in live field soil at the 200 mL (Hedges’ g: -1.74; 95% CI: -2.46 to -0.98) and 400 mL (Hedges’ g: -1.96; 95% CI: -2.80 to -0.98) root additions. In potting soil, S. scoparium biomass was reduced at the 100 mL (Hedges’ g: -1.04; 95% CI: -1.75 to -0.24), 200 mL (Hedges’ g: -1.32; 95% CI: -1.98 to -0.68), and 400 mL (Hedges’ g: -0.91; 95% CI: -1.61 to -0.14) levels of root addition.
We computed a feedback ratio to have a standardized metric that could be used to contrast the observed effect of root treatments across the three soil types (Fig. 3). In potting soil, E. canadensis had significant negative feedback with increased root addition when comparing the control and the 200 mL (t ratio = -2.0, p = 0.04) and 400 mL (t ratio = -2.24, p = 0.02) feedback ratios. Negative feedbacks were observed for S. scoparium grown in potting soil when contrasting the control and the 200 mL (t ratio = -2.15 p = 0.03) feedback ratios. In live field soils, B. inermis had strong negative feedbacks at all root-addition levels (t ratio < -3.0, p < 0.01). Elymus canadensis grown in live field soil had reduced growth when contrasting the feedback ratios of the control and 400 mL root-addition treatments (t ratio = -4.49, p < 0.001). Suppression of growth was observed for B. curtipendula when contrasting the feedback ratios of the control and the 200mL (t ratio = -3.01, p < 0.01) and 400 mL (t ratio = -3.82, p < 0.001) treatment groups in live soils. Similarly, S. scoparium had reduced growth in live soils for the contrasts of the control and the 200 mL (t ratio = -2.41, p = 0.02) and 400 mL (t ratio =-2.94, p < 0.01) root treatments. Sterile field soils had a mix of positive and negative feedbacks. Contrasts between control and the 200 mL (t ratio = -2.06, p = 0.04) and 400 mL (t ratio = -3.04, p < 0.003) root treatments showed negative feedback for E. Canadensis growth in sterile field soils. Conversely, a net gain in biomass was shown in the contrasts of feedback ratios at all levels of root treatment for B. curtipendula (t ratio > 2.3, p < 0.02) and S. scoparium (t ratio > 7.7, p < 0.001).
Within each soil type, we evaluated the overall proportion of biomass of each species relative to the total biomass in each pot at each level of root addition treatment. The proportion of B. inermis biomass did not change under root treatments in live or sterile field soil. However, when comparing live and field soils, the proportion of B. inermis biomass was reduced in the
0 mL (t ratio = -3.15, p = 0.047), 100 mL (t ratio = -5.44, p < 0.001), 200 mL (t ratio = -4.64, p < 0.001), and 400 mL (t ratio = -4.37, p = 0.001) levels of root treatment (Fig. 4a). No difference in proportion of biomass was detected for E. canadensis at each level of root treatment within or between live or sterile field soil types (Fig. 4b). Within live and sterile field soils, the proportion of B. curtipendula biomass did not differ at any level of root treatment. However, the proportion of B. curtipendula biomass increased in live compared to sterilized soil at the 0 mL (t ratio = 4.74, p < 0.001) and 100 mL (t ratio = 3.76, p < 0.01) root-treatment levels (Fig. 4c). No difference was observed in the proportion of S. scoparium within live or sterile soils at each level of root treatment. We found an increase in the proportion of S. scoparium biomass from the live soil group when compared to sterile soils at the 0 mL (t ratio = 5.68, p < 0.001), 100 mL (t ratio = 5.43, p < 0.001), and 400 mL (t ratio = 6.46, p < 0.001) levels of root treatment (Fig. 4d).