The authors present an interesting and novel study demonstrating the potential for the Skewness-Kurtosis relationship to explain and predict assembly rules in plant communities. While I believe this could be an important contribution to community ecology, I have a number of comments on the current manuscript.

1) Throughout my reading of the manuscript, it was clear that skewness and kurtosis are mathematically related – increasing skewness tends to make a relationship more peaky, and thus will increase kurtosis (thought the two need not always be related). The authors do acknowledge this relationship (lines 364-373). However, they only recommend that some caution should be used in the interpretation of this relationship, or some other ways to interpret the relationship. I believe the authors need to go further than this. The SKR is the fundamental aspect of the entire study, and the authors should address the consequences (if any) of the relationship between skewness and kurtosis.

2) The authors never really address what traits are best used the model. Presumably, this is best applied to key functional traits that define resource use and ecological strategies. However, It would be useful to include some interpretation of how real traits would fit within this framework. Many widely measured functional traits, such as specific leaf area, can have complex relationship with ecological strategies, and this may make stochastic processes more important (or overwhelm the deterministic processes). In addition, some functional traits are not independent dispersal traits (for example, plant height it both a functional trait and a correlate of dispersal ability) – thus the choice of trait will have potentially serious implications on the effectiveness of the SKR to predict community assembly.

3) The overlap between scenarios are used to determine the capacity to differentiate these scenarios. However, no indication of how the authors determined the overlap to be sufficiently low enough to conclude the scenarios were actually different. The authors should include some measure of statistical significance here. The ability to assess the significance of these relationships will be important for others to use the SKR to assess predictions of community assembly in real communities (where the differences in overlap may not be as clear).

4) Line 139 – what underlying processes are being referred to here?

Following my earlier review, I believe the authors present an interesting and potentially powerful approach to understand trait distributions in plant assemblages. I believe the authors made appropriate responses to the previous review. I have some additional comments on the revised manuscript.

1) Trait distributions in community assemblages can be thought of in a similar way to fitness functions in evolutionary biology. However, there is a limit to these similarities. For example, unlike in a fitness function, the peak value of trait distribution does not necessarily make it an optimum value (though it may usually be an optimal value). Following the example of asymmetric competition for light used by the authors throughout the manuscript, an assemblage of plants may include individuals of the tallest (and due to the nature of light), optimal value for height. There may also be numerous other shorter species living in the understory (suboptimal but common in the community). This community may be represented by few optimal individuals and many suboptimal individuals. I don’t think the model presented can differentiate between optimal and dominant trait states.

In addition, unlike fitness functions, we can’t be sure that the trait used in the simulations, or in a study is the main trait driving controlling assembly processes. I suggest the authors limit their interpretation of optimal trait expression.

2) Line 62-63 - The opening statement of the introduction could be clarified – what are the deterministic processes impacting species assemblages?

3) Line 64-65 (and throughout the manuscript) – trait diversity reflects niche differentiation (and other deterministic processes) if that trait is central to (or at least relevant) to these processes.

4) Line 130 – It would be useful to include more information on these trait based scenarios here.

5) Line 323-327 – This may be a useful tool to explore assembly processes. However, different processes can yield the same or similar SKR relationships. This suggests that SKR relationships may not be on their own used to understand assembly processes.

6) Line 333 – “consistent”

Following my earlier reviews, I believe this study makes a potentially important contribution in understanding community assembly and the diversity of traits within assemblages. The authors have done a good job in responding to the comments from previous reviews. I have only a small number of minor comments.

1) Line 322-323: “Therefore, our study therefore demonstrates …” – delete one therefore.

2) Line 327: The current wording is a bit difficult. Change to “effectively discriminate”?

3) Line 334-335 – Is there a reference that could be used to demonstrate our current focus has been on differences in trait distributions?

4) Line 394 – I don’t think it is generally correct to end a paragraph with a colon.