Lines 376-378: I don't see the direct link between homogenization and equality of change among losers and winners. Either prove/show the link mathematically, or drop the statement.Probably, this disagreement is based on how we define homogeneity. In our opinion, homogenization is the direct consequence of redistribution of the species' cover. If decreases are distributed more equally (that is also more homogeneously) across many species and increases in cover are concentrated in few species, the latter (that is the winners) will be increasing in many communities. In consequence, the dissimilarity in species composition between these communities has to decrease (given that a quantitative dissimilarity measure is used). Mathematically, this would have to be shown by a decrease in dissimilarity, which however, is difficult to demonstrate across all plot records in our data set as many communities have no species in common. In our opinion, homogenization probably occurs within habitat types, but opening this discussion and carrying out the analysis would open a can of worms. Thus, we have decided to down-tune this statement to: "Homogenisation occurs because, across all time series, few species consistently increase in their cover, meaning that the same species are winning in many communities." (new l 372-374)Referee #3 (Remarks to the Author):As before, I commend the authors on their analyses and believe this paper makes a novel and important contribution by documenting long-term plant biodiversity changes in terms of cover that would have been missed by simply focusing on species richness, as most previous work has done. But while the authors have produced a strong revision of their paper, in my opinion several outstanding issues remain, and some important comments have only been partially addressed. I appreciate the new analyses that the authors have performed, and overall the analyses are appropriate and justified, and data are presented correctly, as far as I can judge. The framing of the study is also now more compelling, and I think the unique value of the dataset and insights arising from it are now harnessed more effectively. That said, I still have some concerns about the framing of the study, explained in my comments below. The Results section has also been improved, but remains difficult to read in places -again, I make specific suggestions below. Overall, while the length appears to have been reduced, the text still seems unnecessarily wordy in places (e.g. in the Results and figure captions).Thank you for this positive assessment.Another aspect I raised previously and has still not been resolved, in my opinion, is when results are considered to be ecologically relevant or not, and I think there could be greater transparency in the paper about this. This relates to the way the story is set-up, as mentioned in my previous review. Specifically, the authors note that due to the large sample sizes, changes in species richness can be statistically significant even if effect sizes are small, and therefore conc...
