The manuscript by Neri et al. explored how maximum PSII yield changes with PFT and climate using data collected from the literature. The research topic was of great importance for the global carbon cycle, and implementing the idea in terrestrial biosphere models will help improve the model predictions. The manuscript was overall well written, and ideas were well-delivered. While I am convinced about the importance of the idea, I have some concerns about the research and analyses performed. Below are two primary issues I found, and I hope they are useful for the authors.

1. Simply modifying Phi_PSIImax is not adequate for photosynthesis and thus fluorescence models. For example, if the change of Phi_PSIImax is due to those of the rate constants, such as Kd, Kf, Kn, and Kpmax, prescribing Phi_PSIImax will only impact the calculation of electron transport rate J and thus Aj and Agross. However, the subsequent qL, NPQ, and Phi_f calculations will not be accurate as the Kd/f/n/pmax are not changing accordingly. Therefore, a more process-focused model to explain Phi_PSIImax will be more useful. For example, the van der Tol et al. (2013) fluorescence model assumed that Kd is temperature-dependent to explain the temperature dependency of Phi_f on temperature. A similar approach, such as a revised Kn(temperature) function, can be taken here.
2. The authors did not distinguish “response” and “acclimation” in the analyses. For example, let us again assume Phi_PSIImax change is due to those of Kd, Kf, Kn, and Kpmax here. If Kd = a1*T + b1 for plants grown in the C1 environment and Kd = a2*T + b2 for plants grown in the C2 environment, the function a*T + b is “response” (related to temporary changes in the environment), and shift from a1*x + b1 to a2*x + b2 is “acclimation” (related to long term changes in climate). Therefore, it is likely that the data analyzed is a mixture of “response” and “acclimation”, and attributing all the changes in Phi_PSIImax is inappropriate. Without distinguishing the two, the analyses performed might be biased.

Neri et al synthesized PAM measurements of ΦPSIImax from literature, and investigated its temperature responses. A model with interpretable parameters is developed, and a tolerance-resilience trade-off is identified. The impacts of PFT and climatological temperature on ΦPSIImax tolerance and resilience are also investigated. While plenty ΦPSIImax measurements can be found in literature, a synthesis analysis as this work is absent. The presented work could be valuable to the community by facilitating our understanding of photosynthesis temperature response, and providing information for model parameterizations. The manuscript effectively presented the methods and results in general. Below lists my several concerns and suggestions.

1. It is not clear to me how the confounding variables (water, light, etc.) were controlled, although that is stated as a selection criterion (L125). Did you select studies where the confounding variables were controlled in that specific study? My understanding is those variables can still vary from one study to another, and may play a role in the analyses. Could you clarify this?
2. PFT-specific CTI and percent prediction explained (Eq.7 and Figure 5): My understanding is that the PFT-specific CTI is still one equation generated for all PFTs, rather than one equation for each PFT. Is this correct? Could you explain the reason for using a general equation instead of one equation for each PFT? Presenting the values of the aL parameters might also be helpful.
3. Rearranging Section 3.2 and Section 3.3 and putting the CTI map (Figure 8c) before Figures 6 and 7 may help the audience more easily interpret results related to CTI.
4. Are the CTIs in the result section the general CTI?
5. The manuscript is quite long, I suggest cutting the length of the manuscript. Some method and result sections could potentially be moved to the supplementary. For example, details of ART ANOVA, section 3.2.1, and section 3.2.3.