Uniform regulation of stomatal closure across temperate tree species to sustain nocturnal turgor and growth

Abstract

Water loss and carbon gain are balanced by stomatal control¹, a trade-off that has allowed trees to survive and thrive under fluctuating environmental conditions^2-4. During periods of lower water availability, stomatal closure prevents excess water loss⁵. Various strategies of stomatal control have been found among tree species^6,7, but the trigger for this behaviour remains elusive. We found a uniform pre-dawn water potential threshold (-1.2 MPa) for stomatal closure across species, which coincided with stem-growth cessation. Meanwhile, midday water potentials at stomatal closure were more variable across species and stomatal control did not follow species-specific thresholds of hydraulic failure, a commonly adopted theory in plant biology^8-10, and often used in predictive water-use modelling^11,12. This indicates that nocturnal rehydration, rather than daytime hydraulic safety is an optimization priority for stomatal closure in trees¹³. We suggest that these processes are critical for forecasting the global carbon cycle dynamics.

Fig. 1. Time series of stomatal conductance (g_s) and pre-dawn leaf water potentials (Ψ_leaf) collected using a canopy crane at the Swiss Canopy Crane II (SCCII) site.

a, Mean species-specific midday g_s measured on leaves from the top of the canopy, shown with large, coloured circles. Open circles indicate measurement dates where Ψ_leaf was below −1.2 MPa. Grey dots represent raw measurements. b, Pre-dawn (04:00–06:00 CET) Ψ_leaf, with coloured dots indicating the species mean. In 2022 an exceptional drought caused all Ψ_leaf to drop below −1.2 MPa (shown with open circles). c, Theoretical example of how the point of stomatal closure (P_st) was defined. P_st was assumed when the negative exponential behaviour of g_s changed to a linear decrease (Supplementary Methods). Midday Ψ_leaf values are presented in Supplementary Fig. 3. J through D, months of January through December. Source data

Fig. 2. Relationship between pre-dawn and midday Ψ_leaf and normalized midday g_s or growth of mature temperate tree species.

a, The relationship between normalized midday g_s and pre-dawn Ψ_leaf for all studied species at the SCCII site. The point of stomatal closure (P_st) was quantified for each species individually. The across-species mean P_st is indicated by a bold line, with the 50th and 95th percentile ranges shown as grey shading (Supplementary Fig. 4). b, Instantaneous response conditions between midday g_s and midday Ψ_leaf, presented as in a. The across-species mean, as well as the 50th and 95th percentile ranges for P_st, are depicted as for a. c, Growth probability and rate responses to pre-dawn Ψ_leaf, obtained from weekly band dendrometer readings. Zero growth rates versus non-zero growth rates were analysed to determine the growth probability using a mixed-effect model with a binomial distribution. Axes labelled (-) indicate unitless values, resulting from normalization to the maximum value. The grey shading around the line represents the 95% confidence interval. Source data

Fig. 3. Sap flow and g_s response to pre-dawn Ψ_leaf across broad spatial scales.

a, Pre-dawn Ψ_leaf measurements related to daily maximum sap flux density (expressed as a percentage of the maximum sap flux density of the tree). Species are distinguished by different colours. The line represents a significant log-transformed relationship, with species considered as a random effect to generate an across-species response curve. The grey dotted line is shown at −1.2 MPa for reference. b, Locations of the sites included in this analysis, as detailed in Supplementary Table 5. c,d, Relationship between pre-dawn Ψ_leaf and g_s for Australian tree species (c), distinguishable by colours as presented in the location map (d). Source data