Evaluating Native Grassland Species for Application in Extensive Green Roofs in Japan^†

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

In the present study, titled “Evaluating Native Grassland Species for Application in Extensive Green Roofs in Japan,” the authors evaluated six native Japanese grassland species on extensive green roofs across three substrate types (reused, perlite-reused mix, and perlite-based) over one growing season. Plant cover, flowering, and survival were measured, with a focus on the feasibility of reusing old green-roof substrate. Overall, the paper is interesting and well-written, and it could contribute to the existing knowledge base on using native species in extensive green roofs. The main drawback is the short study period, which limits strong conclusions, especially because the period did not include pronounced drought stress. Nevertheless, the manuscript fits the scope of Environments and could be suitable after revision.

 

Specific comments:

Please state the exact months and year of the study in the Methods and in the Abstract. Also, mention the exact sampling dates in the Methods.

The Methods mention twelve 1 × 1 m quadrats (4 replicates × 3 substrates). In the statistical section, however, the site is divided into 40 blocks (line 203). This is potentially misleading and needs clarification.

Substrate descriptions need more detail. The reused substrate was sieved through a 6 mm sieve, which is quite small. Please report its original composition or at least its constituents. For the perlite-based substrate, provide the particle-size distribution and whether any organic fraction was present. If I understand correctly, is this a perlite-only substrate? In the green roof literature and standards, perlite is typically a component of substrate mixes, not the sole medium. FLL guidelines indicate extensive green roof media is 80-90% mineral with approximately 10% organic matter by volume. Using pure perlite is atypical outside a hydroponic context. Could the authors cite examples of perlite-only extensive green roof media? Perlite alone is nutrient-free, easily washes away, may not retain enough water during dry periods, can lack structural stability, etc.

Figure 4 appears to show only one plant of I. cylindrica per quadrat. If so, please state and justify that.

The current, highly detailed equation-by-equation presentation of the statistical model makes it hard to follow. I recommend condensing the main text into a short paragraph stating what was modeled and why, and moving the equations and other details to a Supplement. The authors could provide the raw data in the Supplement as well.

Improve clarity of cover figures. Spell out substrate codes in the legend.

Author Response

The responses to Reviewer 1's comments are provided in the uploaded PDF file.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This paper reports on a straightforward rooftop horticultural trial for six native Japanese grassland species. The experiment was conducted in a shallow soil green roof system (extensive green roof), over one year. The authors recycled the structure and substrate from a previously installed green roof system on the same roof. They tested reused substrate alongside a mixture of reused substrate and perlite and a perlite-alone treatment. The substrate depth was 100mm.  Species choice was based in part on observations of spontaneous establishment of native species on green roofs originally planted with Sedums, and also the habitat template idea, which suggests using vegetation types naturally occurring in exposed, drought-prone condictions (that resemble green roof environments).

They sampled 12 quadrats total in a randomized block design; each planted individual was monitored for survival, and percent cover (again of each plant) was quantified using image analysis. Plant growth and survival was modelled using a Bayesian approach. This model appears to be overly complicated relative to the small dataset, consisting of repeated measures plant survival and size, and especially given that most species had very low mortality. While I think there is no problem with the approach, I would like to see some justification for this, why not use a more straightforward frequentist or model selection approach? The patterns e.g. Figure 6, are clear from the graphs.

The paper is well organized and clearly written. Results are clear and discussion restricted to actual results provided by the authors. While the design prevented evaluation of ecosystem services attributable to each species (all species were planted in each replicate), the inclusion of flowering phenology data yields some intriguing results (to this reviewer), but which the authors did not mention. It looks like there is phenological complementarity among the species for flowering (peak flowering time was different by species e.g. 7c, 9c, 10c, 11c); I recommend that the authors mention this as a beneficial property of including these different species on a green roof: the overall period of flowering is long if all the species are included; any individual species alone would result in much less time of the growing season where flowers are available to pollinators.

Details

ln 172  What were the most common species of weeds that were removed?

ln 570  citation missing journal title (Urban Habitats?)

Author Response

The response to the reviewer 2's comments is provided in the uploaded PDF file.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have adequately addressed the comments. The revised manuscript has improved and can be accepted for publication in its current form.