|After going through the revisions of "The effect of rainfall amount and timing on annual transpiration in a grazed savanna grassland", I find that the uncertainties and limitations to the study have still not been adequately addressed. In particular, estimation of ecosystem transpiration from all methods is difficult, and methods relying on eddy covariance methods, while improving in recent years, are still uncertain (Stoy et al 2019, Nelson et al 2020, Scott et al 2020, Hu and Lei 2021). In depth analysis from sites with many quality measurements and expert knowledge (such as what is presented in this manuscript) are vital to understanding both the true ecosystem transpiration dynamics, as well as uncertainties in the transpiration estimation methodologies. Therefore, utilization of multiple methods for estimating transpiration with different underlying assumptions is important to understand if the patters observed, such as the findings here that annual T and T/ET are linearly related to the early season precipitation, are robust.|
The manuscript uses three ET partitioning methods, with the Berkelhammer and uWUE methods being very similar both in calculation and underlying assumptions. The third method (TEA), also shares many assumptions (particularly that T=ET during some periods), but is the most different of the three. The previous version dismissed the uWUE and TEA methods stating that " The T/ET values in the late wet season of 2015 based on the TEA and uWUE methods are likely overestimates, given the decrease in GPP and low EVI values during this drought year", and did not present findings based on the other methods in many of the results, making it difficult to understand the robustness of the findings and conclusions. I think it is very important to show all the results, even if in the supplementary materials, particularly Figure 7 which is the main finding. However, while all three methods are now presented in many of the plots, the TEA method is still excluded in the key findings.
The argumentation used for dismissing the TEA method is now based on a soil evaporation model, with the derived parameter of De for the Berkelhammer method being closes to those reported in other literature, particularly Hu and Lei, 2021. However, the Hu and Lei study also compared seven different ET partitioning methods (including the uWUE and TEA methods) and found that the TEA method performed the best, which is a direct contradiction of what is reported here.
The issue here is not which method is correct or incorrect, because all the partitioning methods are all wrong in some way. I strongly advise the authors to revise the results here to take this uncertainty into account and understand how the differences in partitioning methods may impact the interpretation, or at the very least report the findings from all the methods.
Hu, Xingyu, and Huimin Lei. “Evapotranspiration Partitioning and Its Interannual Variability over a Winter Wheat-Summer Maize Rotation System in the North China Plain.” Agricultural and Forest Meteorology 310 (November 2021): 108635. https://doi.org/10.1016/j.agrformet.2021.108635.
Nelson, Jacob A., Oscar Pérez‐Priego, Sha Zhou, Rafael Poyatos, Yao Zhang, Peter D. Blanken, Teresa E. Gimeno, et al. “Ecosystem Transpiration and Evaporation: Insights from Three Water Flux Partitioning Methods across FLUXNET Sites.” Global Change Biology, October 6, 2020. https://doi.org/10.1111/gcb.15314.
Scott, Russell L., John F. Knowles, Jacob A. Nelson, Pierre Gentine, Xi Li, Greg Barron-Gafford, Ross Bryant, and Joel A. Biederman. “Water Availability Impacts on Evapotranspiration Partitioning.” Agricultural and Forest Meteorology, November 2020, 108251. https://doi.org/10.1016/j.agrformet.2020.108251.
Stoy, Paul C., Tarek S. El-Madany, Joshua B. Fisher, Pierre Gentine, Tobias Gerken, Stephen P. Good, Anne Klosterhalfen, et al. “Reviews and Syntheses: Turning the Challenges of Partitioning Ecosystem Evaporation and Transpiration into Opportunities.” Biogeosciences 16, no. 19 (October 1, 2019): 3747–75. https://doi.org/10.5194/bg-16-3747-2019.