Most terrestrial plant species form symbioses with arbuscular mycorrhizal fungi (AMF). However, the carbon (C) transferred from plants and the nutritional and growth benefits they receive from AMF vary greatly across species and environments. Currently, the extent to which this variation is governed by plant functional trait syndromes remains poorly understood. To address this, we conducted a ¹³C pulse labelling study with four grass species inoculated with three AMF species under four precipitation regimes to test whether plant C allocation and AMF-derived benefits can be explained by functional traits representing specific axes of the plant economics spectrum (PES). Our results demonstrate that the two main dimensions of the PES differentially regulate the plant–AMF interaction. The first dimension (PC1), strongly aligning with conservation-acquisition strategy, was a predominant predictor of belowground C allocation and its sensitivity to precipitation. The second dimension (PC2), defined by variation in specific root length (SRL) that strongly reflecting a collaboration strategy, primarily governed nutrient-related mycorrhizal responses and their reaction to altered precipitation. Furthermore, path analyses revealed that these traits exert direct and AMF community-mediated indirect effects on the symbiosis. Synthesis: Our results highlight that the positioning along plant economic strategies provides a predictive framework for plant C-allocation and mycorrhizal responses. By demonstrating that the trait-symbiosis coupling is modulated by precipitation, this study advances our ability to predict how plant–AMF interactions respond to environmental changes.

1. Greenhouse experiment data

(1) Seeds collection in the field:

The mature seeds of four common grass species from southern China were collected. Seeds of Chloris barbata Sw. were collected in Macau Special Administrative Region (SAR), P.R. China (22°08′15.32″N, 113°33′29.57″E), while the seeds of Eragrostis tenella (L.) P.Beauv. ex Roem. & Schult., Leptochloa panicea (Retz.) Ohwi, and Oplismenus compositus (L.) P.Beauv. were collected in Guangzhou, Guangdong, P.R. China (23°05′40.98″N, 113°21′17.91″E), in 2020.

(2) Greenhouse experiment:

The experimental design encompassed three treatment factors: plant species (with four species), AMF inoculation (present or absent) and simulated precipitation levels (400, 800, 1200, and 1600 mm). Each treatment combination was replicated five times, resulting in 160 plants. Field-collected seeds of each study species were surface sterilised with 0.5% KMnO₄ solution, germinated in Petri dishes for 14 d, and then transplanted as individual seedlings into 1-L pots, respectively. Each pot was filled with 1000 g substrate and watered until saturated (100% water holding capacity). AMF addition was manipulated by inoculating a 30 g soil inoculum of AMF species, Entrophospora etunicata (Ent), Funneliformis mosseae (Fun), and Paraglomus occultum (Par). Daily doses of 9, 18, 27, and 36 ml distilled water were supplied to corresponding pots to mimic mean annual precipitation (MAP) levels of 400, 800, 1200, and 1600 mm, which lasted for nine weeks starting three weeks post-planting. Additionally, they received a weekly supply of 9 ml of Hoagland nutrient solution three weeks after transplanting. To ensure that each pot experienced similar environmental conditions, all pots were randomly arranged and rotated weekly. The experiment was conducted in a greenhouse over 12 weeks from January to April 2021, at Sun Yat-sen University, Guangzhou, Guangdong, P.R. China (23°04′38.73″N, 113°18′27.29″E).

(3) ¹³CO₂ stable isotope labelling:

To test carbon allocation to shoots, roots, and rhizosphere soil, pots with AMF inocula were pulse labelled with ¹³CO₂ at dawn one day before harvest. The pots were placed into airtight acrylic chambers (each 0.1 m³) equipped with a fan to circulate the inner atmosphere (Slavikova et al., 2017; Wang et al., 2021). The air in each chamber was enriched with ¹³CO_2, reaching approximately 1500 ppm, and then the CO₂ concentration inside the chamber dropped to its initial level. Following the CO₂ labelling, the pots were maintained in the chamber under the same growth conditions until harvest.

(4) Harvesting and measurements:

Plants were harvested nine weeks after experimental precipitation treatments were initiated and separated into roots and shoots. Rhizosphere soil was collected and stored at -20℃ until DNA extraction. Roots were carefully washed free from soil with tap water. Then, all the shoots and roots were dried at 70°C for 72 h, weighed, and ground. Shoot carbon (C) and nitrogen (N) concentrations were analysed by dry combustion using an elemental analyser (vario EL cube, Elementar Co. Ltd, Germany). Shoot samples were digested with 10% v/v ultrapure nitric acid using a high-performance microwave reactor (UltraClave, Milestone, Italy), and P concentration analysis was performed on an ICP-AES spectroscopy (iCAP 6500 Duo, Thermo Fisher Scientific Inc., Waltham, MA, USA).

To determine carbon allocation changes, the isotopic signature of shoots, roots and soil carbon (δ¹³C) for dried subsamples were assessed through the Combustion Module-Cavity Ring Down Spectroscopy (CM-CRDS) system (Picarro, Inc., Santa Clara, CA, U.S.) (Balslev-Clausen et al., 2013). The samples were acidified with HCl 1 N at 40°C until carbonates were completely dissolved and then rinsed with Milli-Q water to remove any acid residues. After undergoing freeze-drying, approximately 50 mg of the samples were analyzed in a small 5 × 9 mm capsule. Calibration was carried out using UREA-¹³C. The precision was typically within 0.03‰ for triplicate analyses.

(5) DNA extraction, sequencing and bioinformatics:

To determine AMF community structure, we extracted soil DNA using the MagaBio® Soil DNA extraction kit and Feces Genomic DNA Purification Kit (0.25 g soil per extraction) (Hangzhou Bioer Technology, Hangzhou, China). We followed the protocols for the quality verification implemented by NanoDrop One (Thermo Fisher Scientific, MA, USA). Then, the purified DNA underwent PCR amplifications using AMF-specific small-subunit (SSU) ribosomal RNA gene primers, AMV4.5NF – AMDGR (Sato et al., 2005; Van Geel et al., 2014). Sequencing was conducted using the Illumina Nova 6000 platform (Guangdong Magigene Biotechnology Co., Ltd. Guangzhou, China). Bioinformatics analysis involved filtering out low-quality sequences (average score <20) and short sequences (<100 bp). UPARSE was used to cluster sequences into OTUs at a 97% similarity threshold for microbial diversity analysis. The resulting identities were analysed for their distribution within the phylum Glomeromycota.

(6) Calculation of proportional ¹³C allocation to aboveground and belowground:

The proportion of ¹³C transferred to the aboveground (shoot) and belowground (root and rhizosphere soil) was calculated using the ratio between the measured ¹³C amount in the shoot, root and rhizosphere soil per pot and the whole of these three as Eqn 1:

% C_x = X_13C / (shoot_13C +root_13C + rhizosphere soil_13C) × 100

Where X stands for the carbon allocation part: aboveground (Ag) or belowground (Bg, root and rhizosphere soi). The values of the rhizosphere soil were corrected using the background δ13C values measured in the non-labelled soil without any AMF inoculum after harvest.

(7) Calculation of mycorrhizal response (MRs):

To assess the impacts of precipitation on plant responses to AMF, we quantified various mycorrhizal nitrogen responses (shoot concentration [MNR], and shoot content [MTNR]), phosphorous responses (shoot concentration [MPR], and shoot content [MTPR]) and growth responses (Shoot_MGR, Root_MGR, Total_MGR) of each individual at each of the four precipitation levels, respectively, using the formula, following the concept of Cavagnaro et al. (2023) (Eqn 2):

% MXR = (X_AM - X_NM )/ X_NM  × 100

Where X stands for the individual variable of MR, which can be either P (P concentration of the shoot), TP (P content of the shoot), N (N concentration of the shoot), TN (N content of the shoot), or G (plant growth, i.e. shoot, root or total biomass). X_NM represents the mean value of X for the non-mycorrhizal treatment.

2. Leaf and root functional traits quantifying data

The four grass species seeds were germinated by the same method as described in the greenhouse experiment before and cultivated in autoclave-sterilized sandy soil in a 0.4-L pot (diameter: 5.50cm, height: 17.20cm) within a controlled climate chamber environment (day: 24°C, 16 h; night: 20°C, 8 h, RV 70%) for ~2.5 months (until seedlings were large enough to harvest). Seedlings were watered to saturation daily with sterile distilled water and a 20ml Hoagland nutrient solution each week to prevent water limitations.

To quantify the functional traits of leaves and roots from those cultivated grass species that are related to the plant economic spectrum, we measured 10 morphological and chemical traits of leaves and roots, including specific leaf area (SLA; cm² mg⁻¹), leaf dry matter content (LDMC; mg mg⁻¹), leaf carbon concentration (LCC; mg g⁻¹) and leaf nitrogen concentration (LNC mg g⁻¹), specific root length (SRL; cm mg⁻¹), specific root area (SRA; cm² mg⁻¹), root diameter (RD; mm), root tissue density (RTD; root DW per volume mg cm⁻³), root carbon concentration (RCC; mg g⁻¹) and root nitrogen concentration (RNC; mg g⁻¹). were measured. For each species, measurements were taken from five individuals to calculate leaf and root traits.

Three young, fully expanded leaves (only lamina is considered) of each individual were selected. Five punches of each leaf were collected randomly, and their fresh mass was measured. All punches were then dried at 60°C for 72 h, weighed again, ground and subsampled for elemental analysis. Leaf dry matter content (LDMC) was quantified using the fresh and dry masses and specific leaf area (SLA) was quantified using the punch area and dry masses. Leaf carbon concentration (LCC) and leaf nitrogen concentration (LNC) were measured using the same method as described in the greenhouse experiment.

The roots of each seedling were carefully washed free from the sand and stored in a 4°C refrigerator until morphological measurement. For each seedling, the average SRL and diameter of absorptive root branches were measured using images acquired with a ScanMaker i850 scanner (MRS-9600TFU2L; Shanghai Microtek Technology Co., Ltd, Shanghai, P. R. China) and analyzed using LA-S Plant Root Analysis System (Wan Shen, Hangzhou, P. R. China). Images were acquired using transmissive scanning, at a resolution of 800 dpi, and 16-bit grayscale format. After scanning, roots were dried at 60°C for 72 h, weighed for their dry mass, ground, and subsampled for elemental analysis. Root images were then analyzed using the LA-S Plant Root Analysis System (Wan Shen, Hangzhou, China). The root traits were measured in terms of the total root length, root surface area, root volume, and root average diameter (RD). and then specific root length (SRL) was calculated as root length divided by root dry mass, specific root area (SRA) was calculated as root surface area divided by root dry mass, and root tissue density (RTD) was calculated as root dry mass divided by fresh root volume.

References

Balslev-Clausen D, Dahl TW, Saad N, Rosing MT. 2013. Precise and accurate δ13C analysis of rock samples using Flash Combustion–Cavity Ring Down Laser Spectroscopy. J. Anal. At. Spectrom 28: 516–523.

Cavagnaro TR, Smith FA, Ayling SM, Smith SE. 2003. Growth and phosphorus nutrition of a Paris-type arbuscular mycorrhizal symbiosis. New Phytologist 157(1): 127-134.

Van Geel M, Busschaert P, Honnay O, Lievens B. 2014. Evaluation of six primer pairs targeting the nuclear rRNA operon for characterization of arbuscular mycorrhizal fungal (AMF) communities using 454 pyrosequencing. J Microbiol Methods 106: 93-100.

Sato K, Suyama Y, Saito M, Sugawara K. 2005. A new primer for discrimination of arbuscular mycorrhizal fungi with polymerase chain reaction-denature gradient gel electrophoresis. Grassland Science 51(2): 179-181.

Slavikova R, Puschel D, Janouskova M, Hujslova M, Konvalinkova T, Gryndlerova H, Gryndler M, Weiser M, Jansa J. 2017. Monitoring CO2 emissions to gain a dynamic view of carbon allocation to arbuscular mycorrhizal fungi. Mycorrhiza 27(1): 35-51.

Wang R, Bicharanloo B, Shirvan MB, Cavagnaro TR, Jiang Y, Keitel C, Dijkstra FA. 2021. A novel 13C pulse-labelling method to quantify the contribution of rhizodeposits to soil respiration in a grassland exposed to drought and nitrogen addition. New Phytologist 230(2): 857-866.