• The authors make excellent use of a public, curated time-series database to facilitate this study. A major limitation in interpreting the results is the uneven distribution of data across regions (33 marine biomes v. 10 terrestrial and 5 freshwater) and the gap in data from tropical terrestrial ecosystems. It is feasible that inclusion of tropical terrestrial systems would change the mean trend in local species richness and eliminate the differential deviance between marine and terrestrial biomes from the overall trend. While we understand that including tropical terrestrial ecosystems in the current analysis is probably not possible, we would appreciate more discussion of both the cause and effects of this limitation. Why are data unavailable for tropical terrestrial systems? Have they not been collected, or were they not able to meet the selection criteria for the database? While selection criteria for the database is likely discussed in the publication describing its initiation, it may be helpful to readers to briefly how data was selected for inclusion.