Nitrogen Fertilisation Increases Specific Root Respiration in Ectomycorrhizal but Not in Arbuscular Mycorrhizal Plants: A Meta-Analysis

Abstract

Plants spend a high proportion of their photosynthetically fixed carbon (C) belowground to support mycorrhizal associations in return for nutrients, but this C expenditure may decrease with increased soil nutrient availability. In this study, we assessed how the effects of nitrogen (N) fertiliser on specific root respiration (SRR) varied among mycorrhizal type (Myco type). We conducted a multi-level meta-analysis across 1,600 observations from 32 publications. SRR increased in ectomycorrhizal (ECM) plants with more than 100 kg N ha^-1 applied, did not change in arbuscular mycorrhizal (AM) and non-mycorrhizal (NM) plants, but increased in plants with a dual mycorrhizal association in response to N fertilisation. Our results suggest that high N availability (>100 kg N ha^-1) could disadvantage the growth of ECM plants because of increased C costs associated with maintaining higher root N concentrations, while the insensitivity in SRR by AM plants to N fertilisation may be because AM fungi are more important for phosphorus (P) uptake.

Keywords: association; carbon cost; expenditure; meta-regression; multi-level; multi-model inference; symbiosis; uptake.

Figure 1

Sum of Akaike weights (AICc) from the multi-model inference for predictors including mycorrhizal type (Myco type), N fertilisation levels (N level), soil temperature (Soil temp), gravimetric soil moisture (Soil moist), Specific root respiration measurement method on N fertilisation effect sizes of specific root respiration. Cut-off value (blue line) is set at 0.8 differentiating between important and less important predictors. N.B. only observations from field experiments are included.

Figure 2

Effect sizes (LnRR) of specific root respiration (SRR) to N fertilisation in plants with no mycorrhizal associations (NM), with ectomycorrhizal associations (ECM) only, arbuscular mycorrhizal associations (AM) only, or AM+ECM associations. Numbers in brackets represent the number of observations, and error bars represent the 95% CIs. **indicates significance at p < 0.01. Only observations from field experiments are included.

Figure 3

Effect sizes (LnRR) of SRR in plants with no mycorrhizal associations (NM), with ECM associations only, AM associations only, or AM+ECM associations, for (A) low N fertilisation levels (N ≤ 100 kg ha⁻¹) and (B) high N fertilisation levels (N > 100 kg ha⁻¹). Numbers in brackets represent the number of observations and error bars represent the 95% CIs. ** and * indicate significance at p < 0.01 and p < 0.05, respectively. Only observations from field experiments are included.

Figure 4

Effect sizes (LnRR) of root N concentration to N fertilisation (A) in association with different mycorrhizal symbiosis and (B) meta-regression relating LnRR of SRR to N fertilisation with root N concentration in fertilised treatments (NM, non-mycorrhizal—red symbols; ECM, ectomycorrhizal—blue symbols; and AM, arbuscular mycorrhizal—grey symbols). ***, **, and *indicate significance at p < 0.0001, p < 0.01, and p < 0.05, respectively. Shade band around the regression fit represents 95% CIs. Only observations from field experiments are included.

Figure 5

Conceptual model illustrating the influence of soil N availability on SRR and root N concentration for plants in association with (A) AM and (B) ECM fungi. With increased soil N availability, both ECM and AM plants spend less C to support mycorrhiza resulting in reduced SRR associated with mycorrhizal symbiosis (a dark blue component of SRR) but respire more C due to an increase in root N concentration (a light blue component of SRR). However, the increase in plants C costs associated with higher root N concentration outweighs the reduction in C to support ECM fungi, while it counterbalances the reduction in C to support AM fungi. Furthermore, in AM plants, C costs for mycorrhizal symbiosis may remain relatively high at high soil N availability, because AM symbiosis is more important for P uptake. We refrained from drawing models for NM and AM+ECM plants due to limited data.