Climate change-induced shifts in plant functional traits can profoundly affect the stability of plant communities and ecosystem functions. However, the way in which leaf and root traits coordinate in response to climate extremes remains poorly understood. Here we measured sixteen leaf and root traits of ten dominant plant species from a Tibetan alpine grassland to explore whether extreme warming leads to coupled/decoupled responses of root and leaf traits. We found that extreme warming increased the investment in defensive traits of plant leaves, but this came at the expense of plant growth. Specifically, warming increased the concentration of defensive compounds (tannin) in leaves by 44.4% but decreased leaf mass by 12.3%. More importantly, we found that the changes in leaf and root traits in response to extreme warming were coupled, resulting in both systems adopting more conservative strategies. For instance, extreme warming reduced leaf area and simultaneously increased root tissue density. Trait network analysis further supported this view, showing that extreme warming increased root edge density and decreased leaf modularity, indicating a more integrated network of leaf and root traits to enhance resource utilization efficiency. Our findings suggest that extreme warming may drive a coupled shift toward more conservative traits in both leaves and roots within alpine grasslands, with significant implications for predicting future ecosystem dynamics under escalating climate extremes.

Species culture and laboratory-controlled extreme warming experiment

In August 2020, a survey was conducted in the alpine meadows at Haibei Station to assess biomass and identify dominant species based on relative biomass abundance. From May to August 2021, seeds from over 30 common species were collected for germination. Based on survey and pre-experiment results, 10 dominant plant species were selected for the experiment, including 2 types of grasses (Stipa aliena and Elymus nutans), 2 types of sedges (Carex alatauensis and Carex buchananii), 2 types of legumes (Tibetia himalaica and Oxytropis ochrocephala), and 4 types of forbs (Saussurea superba, Saussurea nigrescens, Gentiana straminea and Aster diplostephioides), which together accounted for 48.54% of the total biomass.

In our laboratory-controlled extreme warming experiment, spanning one growing season (from May to September), we employed temperature gradient plates (Model GRD-1; Grant Scientific Inc., Cambridgeshire, UK) to apply both control and warming treatments. The control treatment incubator was set to 16 °C, while the extreme warming treatment was set to 20 °C (+4°C, > 2δ). Each species had three replicates per treatment, with 4-6 plants per pot. In total, 60 pots were used. To minimize the potential impact of pot position on experimental outcomes, the pots were randomly repositioned every ten days throughout the study. To simulate the effects of extreme temperatures on soil moisture, all plants received ample water during the first month of growth. Following this period, watering was reduced to once every 20 days. The light cycle, consisted of 16 hours of light and 8 hours of darkness, simulated the growing season conditions as Haibei Alpine Grassland Ecosystem Research Station.

Plant functional trait measurements

In our study, we monitored 16 morphological and chemical traits. The net photosynthetic rate (shortened as Photo) was measured using a portable Li-6800 photosynthesis system (Li-cor, USA), and chlorophyll content (SPAD) was assessed with a SPAD-502Plus chlorophyll meter (Konica-Minolta, Japan). Root exudates were collected following Phillips et al. (2011) with minor modifications. After selecting healthy plants with intact root systems, removed residual soil and organic matters, then placed the roots in a hydroponic system to collect exudates. The collected liquid was filtered through a 0.45 μm membrane within 2-4 hours and stored at -20 ℃. The root exudation rate (RE, μg C⁻¹ g root biomass h⁻¹) was determined by measuring total organic carbon with a TOC analyzer (Multi N/C 3000, Analytikjena, Germany) using 15 mL of collected liquid.

Next, fine roots (< 2 mm in diameter) and leaves were randomly selected for measurement of morphological traits. Roots were scanned with an Epson Expression 10000XL Scanner (Epson, Japan), and leaves were scanned with a portable Li-3000 leaf area meter (Li-cor, USA). Scanned images were analyzed for leaf area (LA), average diameter of fine root, fine root length, surface area, and fine root volume using WINRHIZO software (Regent Instruments Inc., Canada). Both leaves and roots were dried at 105°C for 30 minutes, then at 75 °C for 48 hours to a constant weight. The dried samples were weighted to calculate specific leaf area (SLA), leaf dry matter content (LM), specific root length (SRL), specific root area (SRA), and root tissue density (RTD).

For measurement of chemic traits, the leaf and root samples were milled. Carbon and nitrogen contents of leaves (LC, LN) and roots (RC, RN) were measured using an elemental analyzer (CHNS/O, Leeman, Italy). Leaf tannin content (Tannin) was determined by shaking 1-2 g of dried leaf sample in an acetone solution for 30-40 minutes, filtered, analyzed absorbance spectrophotometrically, and then calculated from a standard curve where tannic acid concentration served as X-axis and absorbance values as Y-axis. Leaf lignin content (Lignin) was measured using a lignin detection kit, with the spectrophotometer preheated for 30 minutes and set to 280 nm, using acetic acid as the zero reference.

Changes after Jun 19, 2025:

In the previous version of the dataset (Jun 19, 2025), we identified inconsistencies in some variables (LC, LN, LA, LM, SLA, Lignin, RSR) compared to the experimental records. The data has now been corrected based on the experimental records, and a new version has been re-uploaded. An official Erratum has been issued through the journal (https://spj.science.org/doi/10.34133/ehs.0413).