I agree with your responses to the reviewer comments on your paper. Your paper is mainly about a methodology to determine a pedotransfer function using a boosted regression tree and how the parameters of this model could be derived (using grid search or a differential evolution algorithm). In view of this, I think it is very important to address the comments by Alexandre Wadoux and include your responses in your paper.
I would propose to put less focus on the derived pedotransfer functions themselves but more on the procedure on how to derive pedotransfer functions. The title of your paper puts the focus on these pedotransfer functions for tropical mountain soilscapes. But, would your pedotransfer functions be applicable for any other tropical mountain soilscape than the two that you considered? Since you developed two different pedotransfer functions, one for each landscape, I really doubt this. In the paper, these transfer functions are rather black boxes and it is far from clear whether they have any predictive power for other sites, land uses, …, than the sites where they were developed for. They do not provide insights in which input variables are most important to explain the output variables (at least, that was not reported in the paper) and it is not clear how they could be applied to datasets with different input data.
You are going to make the pedotransferfunctions that you developed available. It is however also necessary to make the dataset with input and output data available. This is important for the further improvement and generalization of pedotransfer functions.

Thank you for the revising the paper. I think that all comments were appropriately addressed. I just have one technical issue that needs tp be corrected:Real-valued parameters: this is not correctly used throughout the paper. Real-valued refers to real numbers which are opposed to imaginary numbers. All the parameters you are considering are real valued numbers. Please change real-valued to continuous parameters and put this against discrete parameters that are considered in grid search methods.

Dear authors,

Based on the reports of the TE and the revised ms, Im happy to inform that your ms is now ready for publication. However, the TE has some important technical corrections that need to be implement. Please carefully consider them before submitting your final ms. Thanks you for submitting your work to SOIL.

Kind regards,

K Van Oost