Dear Authors,
You replied to all comments received by SOIL. In some cases the response was a bit general and not stating what exactly you plan to change in the next manuscript version.
Please note that there is still one issue not completely resolved: on the prediction limit. First, if the units are g.kg-1 then is is not a stock but a concentration; second, an upper limit of 2980.95 g.kg-1 means that SOC content can exceed 100% which is not possible. Please take this into account as well as the edits based on the reviews.
I invite you to submit a revised version of the manuscript according to your replies to the comments and will evaluate the revised version again.
Thanks,
Peter Finke

Comments to version 4
No Silver Bullet for Digital Soil Mapping: Country-specific Soil Organic Carbon Estimates across Latin America, by Mario Guevara et al.
The authors have substantially adapted the text and have addressed the issues raised by the reviewers. Unfortunately, not all changes are in correct English, so I have compiled a considerable list of small text unclarities or errors that can in most cases easily be dealt with.
Text comments
• p.4/25: “are also contributing with” > “are also contributing to”
• p.5/15: “We base our methodological approach in” > “We base our methodological approach on”
• same line: “on open sources” > “on open source”
• p.5/19: “This dataset represent” > “This dataset represents”
• p.5/25: “equation because showed less extreme values” > “equation because it showed less extreme values”
• p.6/2: “Because the limitations and uncertainty” > “Because of the limitations and uncertainty”
• p.6/9: “interpolating the values in a country basis” > “interpolating the values on a per country basis”
• p.7/3: “but consider the geographical” > “but considers the geographical”
• p.7/13: “a public source of prediction factors.1” > “a public source of prediction factors.”
• p.7/25: “We predict in a country specific and” > “We predict on a country specific and”
• p.8/16: “10-fold validation strategy” should this not be “10-fold cross-validation strategy”?
• p.9/17+18: “In addition, model independent residuals” > “In addition, independent model residuals” (this is not the same)
• p.9/22: “by increasing (every 10) the maximum prediction limit”: Not clear. What is the dimension of 10? Ton.ha-1?
• p.10: Caption figure 2 is unclear. Does 2B refer to the usage of each of the data points as validation point?
• p.10/6: “with less density of of SOC estimated values” > “with lowest density of SOC estimated values”
• p11/4: “temperature-related-“ > “temperature-related”
• p.13/caption fig.3: “Note how the points are close each other” > “Note how the points are close to each other”
• p.14/3: “0.18, 0.35, 0.32 0.1;” > “0.18, 0.35, 0.32 and 0.1;”
• p.14/24: “r20:39” > “r2 = 0:39”
• p.14/26: “validation show” > “validation shows”
• p.14/27: “uncertainty map from the global model reveal” > “uncertainty map from the global model reveals”
• p.14/29: “higher uncertainty exist” > “higher uncertainty exists”
• p.15/caption fig.4: “Please note that the position of each point/method vary” > “Please note that the position of each point/method varies”
• p.16/figs.5 and 6: Should the legend not be in Tg instead of Ta? Maybe I missed it, but how and where is Ta/ha defined? Ta is no SI-unit. See also in text in section 3.7.
• p.17/8: “we report a similar stocks” > “we report a similar stock”
• p.19/10: “was generated in a country-basis” > “was generated on a country-basis”
• p.19/17: “and results in a” > “and this results in a”
• p.20/13: “mapping purposes (Møller et al., 2018)” > “mapping purposes (Møller et al., 2018).”
• P.21/8: “important consideration selecting an efficient modeling strategy, specially dealing with legacy” > “important consideration when selecting an efficient modeling strategy, especially when dealing with legacy”

Thanks for the improvements and congratulations with the acceptance of the manuscript.
Best regards,
Peter Finke

Dear Authors,
Thank you very much for choosing SOIL as a venue to publish your work and working with the reviewers and editors to improve the manuscript.
It is with great pleasure that I accept your manuscript for publication!
Best,
Jo