Extended Data Fig. 2: Site example and aggregated results for the perturbation method.
From: Reduced water loss rather than increased photosynthesis controls CO2-enhanced water-use efficiency
Panels a-c show the isolated CO2 effects for iWUE, Gc, and GPP at an example site, illustrating the quantification of \(\Delta {{\rm{Y}}}_{{\rm{CO}}2}\) based on actual data. Black scatters and lines represent the control experiment (CTL), while colour-coded scatters and lines represent the mute_CO2 experiment. The scatters represent the annual values. The annual values of iWUE and Gc are obtained by taking the median, while annual GPP is calculated by taking the sum over the growing season. For inter-site comparison, original annual values are normalized by the site-specific mean and represented as percentage changes relative to the site mean. Solid lines represent the long-term trends derived from annual data. The difference in trend slopes between CTL and mute_CO2 experiments is calculated as \(\Delta {{\rm{Y}}}_{{\rm{CO}}2}\). Panel d shows the aggregated \(\Delta {{\rm{Y}}}_{{\rm{CO}}2}\) values across 63 EC sites. Each row corresponds to a target variable (Gc, GPP, and iWUE), and each column represents an EC site. The sites are categorized into three groups, denoted by different font colours, based on their divergent response patterns (the joint-response pattern, Gc-only pattern, and minimal-response pattern). The colourmap for Gc is inverted relative to GPP and iWUE, so that more bluish colours consistently represent stronger CO2 effects.
