Stomatal responses of terrestrial plants to global change

Abstract

Quantifying the stomatal responses of plants to global change factors is crucial for modeling terrestrial carbon and water cycles. Here we synthesize worldwide experimental data to show that stomatal conductance (g_s) decreases with elevated carbon dioxide (CO₂), warming, decreased precipitation, and tropospheric ozone pollution, but increases with increased precipitation and nitrogen (N) deposition. These responses vary with treatment magnitude, plant attributes (ambient g_s, vegetation biomes, and plant functional types), and climate. All two-factor combinations (except warming + N deposition) significantly reduce g_s, and their individual effects are commonly additive but tend to be antagonistic as the effect sizes increased. We further show that rising CO₂ and warming would dominate the future change of plant g_s across biomes. The results of our meta-analysis provide a foundation for understanding and predicting plant g_s across biomes and guiding manipulative experiment designs in a real world where global change factors do not occur in isolation.

Fig. 1. A conceptual diagram depicting the physiological mechanisms underlying the effects of global change factors on stomatal conductance (g_s).

+, –, and ? indicate positive, negative, and uncertain effects, respectively. C_i: intercellular CO₂ concentration, ABA: abscisic acid, ROS: reactive oxygen species, VPD: vapor pressure deficit.

Fig. 2. Responses of g_s to global change factors (GCFs).

a The overall changes of g_s in response to GCFs. b The g_s sensitivities in response to GCFs. $△$ in (b) represents one unit of eCO₂ (100 ppm increase), eT (1 °C increase), iP/dP (10% change), eN (1 g m^–2 yr^–1 increase), or eO₃ (10 ppb increase). The weighted mean values are reported with error bars indicating the 95% confidence interval. All the responses are significantly different from zero at P < 0.05. The numbers outside and inside parentheses represent the number of species (n_sp) and observations (n_ob), respectively. See Fig. 1 for variable abbreviations.

Fig. 3. Stomatal responses to global change factors in relation to treatment magnitude.

a–f Natural log-transformed response ratio (lnRR) of g_s to GCFs. g–l Natural log-transformed sensitivity (lnSens) of g_s to GCFs. The size of each point represents the adjusted weight of each data point, and the darker the color means the higher the point density. The error bands surrounding the regression lines represent the 95% confidence interval. See Fig. 1 for variable abbreviations.

Fig. 4. Stomatal responses to global change factors in relation to plant ambient g_s.

a–f Natural log-transformed response ratio (lnRR) of g_s to GCFs. g–l Natural log-transformed sensitivity (lnSens) of g_s to GCFs. The size of each point represents the adjusted weight of each data point, and the darker the color means the higher the point density. The error bands surrounding the regression lines represent the 95% confidence interval. See Fig. 1 for variable abbreviations.

Fig. 5. Interactive effects between global change factors on g_s.

a The overall changes of g_s in response to two-factor combinations. The error bars represent ±95% of the confidence interval for the weighted means with filled and open points indicating significant (P < 0.05) and insignificant (P > 0.05) differences from zero, respectively. The numbers outside and inside parentheses represent the number of species (n_sp) and observations (n_ob), respectively. b Relations between lnRR (natural log-transformed response ratio) of two-factor combination and sum of lnRR of the corresponding individual factor. The error bands surrounding the regression lines represent the 95% confidence interval. See Fig. 1 for variable abbreviations.

Fig. 6. Predicted changes in g_s across biomes by the end of the twenty-first century under the sustainable emission scenario (SSP1-2.6/RCP2.6).

a Changes induced by elevated CO₂. b Changes induced by warming. c Changes induced by changed precipitation. d Changes induced by elevated N deposition. e Changes induced by elevated O₃. The left panels display global maps depicting changes in g_s, with grey and white land colors indicating areas where changes are statistically insignificant and where is a lack of data, respectively. The right panels depict biome-level predictions, with error bars representing the 95% confidence interval. The numbers outside and inside parentheses represent the number of species (n_sp) and observations (n_ob), respectively. Bor.F: boreal forest, Tem.F: temperate forest, Sub.F: subtropical forest, Trop.F: tropical forest, Tem.G: temperate grassland, Med.W: Mediterranean woodland.