The manuscript presents a new R package - rasterdiv - to calculate diversity indices based on remotely sensed data. The innovation of the R-Package is that known diversity indices can now be calculated easily and conveniently using a moving window approach (as shown in Figure 8). Unfortunately, the authors do not succeed in illustrating the application potential of this new R package in ecological research. While they show different global examples of diversity indices calculated with the package, they fail to discuss their application and to interpret the resulting maps. In my view, a local or landscape scale example (which would allow readers to interpret the maps) would have been much better suited here.
The majority of the manuscript deals with the description of (already existing and well-described) diversity indices and their information theory background. While I totally understand the authors’ intention here, I find this presentation of a new software package not convincing:

My recommendations are therefore as follows:

• The discussion is very long (for this type of paper), but essentially provides the information that I would have liked to read in the introduction (e.g. “Concerning the data being used, spectral diversity measures computed from satellite images represent a valid alternative to class-based land cover maps for investigating landscapes heterogeneity. For instance, a highly fragmented landscape characterised by a mosaic of crops and seminatural forests suffers from oversimplification when investigated through land cover classes”, l. 370ff) -> Use the essential parts of the discussion for the introduction
• Replace the global example with a local or landscape scale example and provide an interpretation of the computed diversity indices visible in these maps
• Shorten the Information Theory-focused sections in the introduction and remove the paragraph about the benefits of Open Source and other R packages for raster data processing (l. 126-142)