I would like to thank the authors for their detailed and thoughtful responses to the reviewer comments. The authors’ arguments, if sufficiently defended and elaborated in the revised version, appear appropriate to address the specific factual and methodological issues raised in the reviews, but these should be evaluated again by reviewers.
There remains a more general issue of appropriately defining the causal argument of the paper. The authors claim, in effect, that climate caused the closure of the Silk Road. That causal claim is poorly defined, and this lack of definition appears to be at the root of the reviewers’ strongest objections, particularly those of the third reviewer.
This problem appears to be a typical instance of communication issues arising in historical climate research. I have written about these problems in a short article co-authored with Qing Pei: http://pastglobalchanges.org/products/pages-magazine/13128 I am not asking the authors to cite this article but rather to consider the advice it contains, particularly its advice on communicating causal claims. This advice includes emphasis on: (1) the contrast set of the causal claim; (2) the nature of causation specified; and (3) the role of the specified causal agent as one element in a set of sufficient conditions for the outcome. Considering those concepts (which are explained in the short article), my best interpretation of the authors’ causal claim is as follows: “The new evidence presented in this article favors the inference that drought—more than other factors frequently cited in the literature—provides the best available explanation for the timing of the Ming empire’s partial closure of Silk Road trade routes.”
This more restrictive causal claim would be much more defensible and more in line with the evidence and arguments presented here. To say that the drought caused the closure, without any further specification, would imply other things that may not be the case. Drought, for instance, cannot explain the initial vulnerability of the region’s ecology or economic links. Nor can drought alone explain why trade was not resumed more swiftly or vigorously in subsequent decades of the Ming dynasty. Yet to say that drought caused the closure – full stop – may imply such things, particularly to historians. The article’s introduction to climate and history (lines 315-321) contributes to the sense that the article intends to make an overly expansive case for the role of climate in human history. The concluding sentence (lines 482-484) also speaks of “fundamental” causes, which is overly broad and assumes agreement about what is “fundamental” to history. I would recommend changing both these passages.
Furthermore, it is important to recognize that the method of this paper is one of inference or abduction. It is reasoning from effects (traces in the physical and written records) to causes (drought and its impact on the environment and trade). Therefore, before presenting the new findings, the paper should be in a position to specify what kinds of effects (i.e., traces discovered in the physical and written record) would increase the probability that drought was a principal causal agent for the partial closure of the Silk Road and which effects would decrease that probability. I am concerned that the reviewers (especially #3) were uncomfortable with the article’s claims because this specification was lacking, and therefore many different results might have been fitted to the drought hypothesis after the fact.
Therefore, in addition to making the extensive revisions indicated in the reviews, I would also request that the authors begin by clearly defining their causal argument linking climatic and societal change. Once that has been done, I believe it will also simplify some of the discussion of climatic impacts that was contested in the reviews.

After considering the reviews, I invite the authors to make further revisions to their manuscript in order to revise their claims with respect to the causal connection between drought and the decline of the overland Silk Road. The manuscript will then be reviewed by another historian to ensure that the claims regarding the impact of the drought are appropriate and supported by the evidence.
This manuscript has three main elements: (1) an examination of new physical records indicating desertification; (2) an examination of historical records; and (3) a causal interpretation of the findings. Of these, I find the first element the most important and original.
Reviewer 1 has focused his/her evaluation on the first element and found it appropriate, apart from some minor revisions suggested in the notes. I ask the authors to respond to those revisions and/or revise the text accordingly.
Reviewer 2 has focused his/her evaluation on the second and third elements and found them flawed. In particular, s/he does not agree that the opening of maritime trade has been seriously advanced as an explanation for the closure of the overland Silk Road, and therefore s/he regards this as a “straw man” argument. Moreover, s/he rejects the manuscript’s central causal inference: that the new evidence for desertification demonstrates that climate caused the decline of the overland Silk Road. The new introduction to section 4.3 does not adequately address this issue.
Normally, I would regard such a strong review as reason to reject a paper. However, as previously stated, the most important and original element of this paper is its new physical evidence for desertification, which the reviewers have not disputed. Moreover, reviewer 2’s rejection of the central causal inference may reflect different use of language and different varieties of causation used in different disciplines, as discussed in the previous editor report.
Nevertheless, because this paper discusses human history, its causal claims should be acceptable to historians. I therefore request that the authors significantly revise those claims in the manuscript so that it can be sent to another historian for review.
One option would be to offer much more evidence and narrative demonstrating that the desertification was a necessary or sufficient condition for the decline of the overland Silk Road. However, I do not believe that this article is the best forum for such a detailed narrative, which may belong in a journal of environmental history. Nor am I confident that the authors currently have the evidence to demonstrate that drought or desertification was a necessary or sufficient condition for the decline of the overland Silk Road.
Therefore, I would advise a second option: to separate the evidence of drought (the validity of which is not presently in dispute) from the causal claim about human history, and additionally to shrink and specify that causal claim. In other words, the manuscript should not start from the position that it is trying to prove that climate caused the decline of the overland Silk Road, particularly without specifying what findings would constitute proof or disproof of this hypothesis. Instead, the introduction may explain that the climate hypothesis for the decline of the overland Silk Road is a reason to study the proxy and historical record for signs of drought and desertification (or lack thereof). Then the discussion section may address whether the findings do or do not support the climate hypothesis.
Based on reviewer 2’s concerns, I would also conclude that the findings currently offer only weak support to that climate hypothesis, because there is a considerable lag between the drought and the decline. There are ways of making sense of that lag, as discussed in the revised manuscript, so the findings are still consistent with such a hypothesis. However, in light of reviewer 2’s concerns, I would hardly regard the findings as proof of the hypothesis. Moreover, at best the timing of the drought may only explain the *timing of* the decline, and probably not the fact that the Silk Road declined at all. Finally, given the limits of the available evidence, climate emerges as only the best currently available explanation for that timing, not a definitive explanation for the decline of the Silk Road. New evidence could well support other explanations. Thus, in the absence of additional evidence or a more complete historical narrative, I ask the authors to reduce their causal claim to a statement that the findings are “consistent with” a climate hypothesis for the decline of the overland Silk Road or else (as suggested in the previous editor comments) something like: “The new evidence presented in this article favors the inference that drought—more than other factors cited in the literature— currently provides the best available explanation for the timing of the decline of Silk Road trade routes.” I have highlighted passages in the manuscript – including the title of the manuscript itself -- that would need to be revised so as not to exaggerate the causal inference. I must emphasize that in light of reviewer 2’s comments I would regard exaggerated causal inferences as a reason to reject the manuscript.
Finally, as noted on the text, unless the authors can demonstrate that previous scholarship has seriously advanced the theory that the revival of maritime trade undermined the overland Silk Road then they should remove discussion of that theory from the manuscript. The discussion of the conflict hypothesis is, in any case, a more original and compelling contribution of this manuscript.

The new reviewer has given the manuscript a positive evaluation and has requested minor revisions. Some of these concern spelling mistakes (e.g., Turfan), which were corrected in some parts of the manuscript but not others. The most important revisions concern the political and social causes of the decline of the Silk Road. The authors should expand this section of the manuscript to include the additional factors identified by the reviewer.
In light of the reviewer’s comment on social and political causes for the decline of the Silk Road, I would also ask for the following minor changes. In line 80: “for the decline of” should be changed to “for the timing of the decline of”. In other words, the political, social, and economic factors already discussed – as well as those that the reviewer has indicated – seem to account for the vulnerability of the Silk Road to disruption by the late 15th century. The timing of the drought appears to account for why that vulnerable institution declined during particular decades of the 16th century and not sooner or later. Therefore, by specifying the causal claim, the authors make a stronger case for the role of climate and the relevance of their finding. As explained later in the paragraph, climatic changes, particularly drought in arid regions, have been identified as “triggers” for societal change. “Trigger” here also carries the meaning of the cause for the timing of a historical outcome rather than a complete explanation for that outcome. Although it would be possible for a drought to be so severe as to constitute a complete explanation for a historical outcome, that does not appear to be the case in this study or in the others referenced in this paragraph.
Lines 247-249: I would recommend revising this sentence for clarity.
In lines 270 and 278: The phrase “warfare” should be changed to “an increase in warfare” or “an increase in the frequency of warfare”. A quantitative analysis like this one can only establish whether the frequency of warfare was correlated with migration or the decline of the Silk Road. In cannot establish whether some particular war was a necessary or sufficient condition for the decline of the Silk Road.
Finally, I would encourage the authors to undertake a review for correct English language usage in those sections that have been revised or added since the first draft. For example: In line 23 in the abstract, “natural factors have not been adequately discussed as social aspects” should be “natural factors have not been as adequately discussed as social aspects” or better “unlike social factors, natural factors have not been adequately addressed”.