I have finished my reading and evaluation on the Manuscript ID ELE-01068-2019 titled “Assessing metacommunity processes through signatures in spatiotemporal turnover of community composition”, submitted to Ecology Letters. The strength of this manuscript relies in the temporal analysis of metacommunity dynamics and the proposition of Path-Analysis based od dissimilarities to analyze the metacommunity dynamics. I enjoyed the reading and did not found any falls that should need major concerns. Perhaps more details on the simulation process would help readers to write their own programing simulation (improve reproducibility details).
Regarding the “real data”, I think that more efforts to explain the data and its results/discussion would help readers to see better the context in which the proposed framework would help them. As they were presented, I had the impression that they are they just to be there, or because someone asked for. I would like to see a discussion and interpretation for results found for invertebrates, fishes, snails, and plants (as the future users of the method will do)
Also, I would like to see a better closing paragraph. Ending this manuscript by discussing simulations possibilities was not a good way to leave the readers, in my opinion.

Minor comments
Regarding the simulations, what about a completely random scenario? It would be important to show that no important correlations would be found in such scenarios (a null scenario).
During the 10,000 time steps simulation the species pool remained 100? Or there were some extinctions?
Which was the patterns in common and rare species at the end of the simulation time? How the balance between rare and common species at the end of the simulation would affect the results? It is widely known that in ordination analysis and resemblance measures just a few species actually contributes to the final pattern or dissimilarity measure.

You stated in the introduction that:
“The neutral paradigm focuses on the interplay of stochasticity and dispersal, in a simplified approach that does not consider differences of local adaptation between species, thus explaining local species coexistence as a purely stochastic process driven by species frequencies at a larger scale and immigration rates (Hubbell 2001).
However, at the description of simulated scenarios at Page 7, lines 21 and 22, you state that the neutral scenario did not had dispersal to neighboring cells. Did you really mean “without dispersal”? If yes, this scenario was totally dependent on initial species allocation to cells and random mortality/extinction? Hubbell model did not had dispersal?

Page 7 line 37: What about maximum and minimum species richness, or other measure of variability?

Which was the community size at the beginning of simulation trials?

Page 7 line 48: delete the first “on”

In my opinion the authors did a good job in addressing my most of my concerns with the original manuscript.