Elevated CO2 alters relative belowground carbon investment for nutrient acquisition in a mature temperate forest

Abstract

Forests are potential carbon (C) sinks that partially offset anthropogenic carbon dioxide (CO₂) emissions via enhanced C assimilation and productivity. However, the question remains whether mature trees will express sufficient plasticity in nutrient acquisition strategies to support enhanced growth under elevated CO₂ (eCO₂). Trees may sustain growth by investing C belowground to enhance nutrient acquisition, e.g., by increasing root absorptive surfaces for greater soil available resource exploration (a "do-it-yourself" strategy) or utilizing C exudation or mycorrhizal associations as priming mechanisms for nutrient acquisition ("outsourcing"). We show that 4 y of eCO₂ (+140 ± 38 ppm; i.e., +35% above ambient) altered the relative belowground C investment strategies of mature oak (Quercus robur L.) in a 180-y-old temperate forest. Fine-root branching frequency increased 73% under eCO₂. Specific root C exudation was enhanced under eCO₂ (63%), particularly outside the peak growing season, and the exudate C to nitrogen (N) ratio was increased (28%). Ectomycorrhizal (ECM) biomass production increased during leaf fall (17%) while ECM turnover increased almost fourfold under eCO₂. The exudate and root metabolome composition were considerably altered during the late growing season under eCO₂. We find, therefore, that a broad suite of nutrient acquisition strategies are upregulated under eCO₂, with dynamic shifting between different outsourcing and do-it-yourself elements at different times of the year. These belowground changes support the increase in net primary productivity observed in this forest, with implications for the role of mature temperate forests in the global carbon sink.

Keywords: ectomycorrhizal fungi; free-air carbon enrichment; relative response; root exudation; root morphology.

Fig. 1.

Annual RR showing the magnitude of changes in mycorrhizal, exudate, and root properties under eCO₂ relative to ambient [RR = ln(R_e/R_a), where R_e is the response under eCO₂, and R_a is the response under aCO₂]; positive values indicate an increase, and negative values indicate a decrease under eCO₂ relative to aCO₂. Dots represent means for n = 3, across four timepoints with pseudoreplication of n = 18 for root exudation and root traits, and n = 15 for ECM. Error bars are ± variance (v). Response ratios refer to overall effects of eCO₂ across the entire study period, and therefore do not indicate seasonal differences between treatments.

Fig. 2.

Fine root traits and exudation of mature oak trees under aCO₂ and eCO₂, showing (A) root branching frequency, (B) SRL, (C) specific root C exudation per unit root mass, (D) areal C exudation rate adjusted for total root biomass in the O horizon, (E) specific root N exudation per unit root mass, (F) areal N exudation rate adjusted for root biomass in the O horizon, and (G) exudate C:N ratio. For parts (A–C, E, and G), data are presented as box plots to reflect distribution of data, boxes denote the 25th and 75th percentiles and median lines are given for n = 18 pseudoreplicates, whiskers indicate values up to 1.5× the interquartile range. For parts (D) and (F), exudation rates on a mass basis were adjusted using root standing biomass on an array basis (n = 5), and values are mean ± SE (n = 3).

Fig. 3.

Ordination plots of untargeted metabolomics profiles of roots and exudates under aCO₂ and eCO₂, based on PCA, showing (A) root metabolome under positive ionization, (B) root metabolome under negative ionization, (C) exudate metabolome under positive ionization, and (D) exudate metabolome under negative ionization, analyzed via LC–MS. Each small, filled, symbol indicates an individual root/exudate, and the cross indicates the mean (n = 18). The ellipses indicate the 95% CI, and Dim1 and Dim2 indicate the variability explained by principal components 1 and 2, respectively. For each plot, the total number of features, and the number significantly (P < 0.05) accumulated or depleted under eCO₂, determined via Kruskal–Wallis tests, are shown.

Fig. 4.

ECM biomass production rate (A) and turnover rate (B) under aCO₂ and eCO₂ in sand ingrowth bags installed for 3 mo (A) and 6 and 12 mo (B). Bold P values indicate a significant difference at the individual time point, determined via a Wilcoxon rank-sum test, and ns indicates not significant. Boxes denote the 25th and 75th percentiles and median lines are given for n = 18 pseudoreplicates, whiskers indicate values up to 1.5× the interquartile range, and dots indicate outliers.