Site-Level Variation in Parrotfish Grazing and Bioerosion as a Function of Species-Specific Feeding Metrics
, Kyle M Morgan 2, Ronan Roche 3, Cassandra E Benkwitt 4 and Nicholas AJ Graham 4Round 1
Reviewer 1 Report
This study describes feeding ecology of nine parrotfish species using bite rates, scar frequency, size and volume to calculate grazing and bioerosion rates at two locations in the Indian Ocean. The authors show there is considerable interspecific variation as well as variation within the broader excavating and scraping functional groups. Feeding rates and bite dimensions are also compared to body size of fish, illustrating the importance of large bodied fish in grazing and bioerosion on coral reefs and that comparable ecological functions can only be achieved when density of fish is high. Finally, the authors examine how grazing and bioerosion rates compare to biomass of parrotfish, demonstrating that biomass may be a good proxy for these processes.
The manuscript is well written and appropriately referenced, analyses are thorough, results well illustrated (though see comments below on some of the figures) and findings appropriately interpreted.
The study illustrates the value of having species level information when calculating rates of grazing and bioerosion and provides species specific information on feeding rates that will improve estimates of these processes in the future. Overall, I think it’s an excellent addition to the special edition that will provide valuable information for understanding herbivory, bioerosion and carbonate budgets on coral reefs. I have only a few comments for the authors to consider below.
L112 feeding rates will change throughout the day. I assume surveys were evenly spread across this time period with no bias towards surveying a particular size or species? Is it possible to illustrate this in supplementary files?
L134-135 The assumption here is that there were no scars in the survey area before the feeding bout?
L166 As study locations are close to the equator I assume season and variation in water temperature are not major issue here when calculating grazing/bioerosion rates. But seasonal and spatial differences in temperature could influence feeding rates (see refs below). How would these factors be considered for calculating grazing/bioerosion at locations at higher latitude?
Ferreira DE, Peret AC, Coutinho R. 1998 Seasonal grazing rates and food processing by tropical herbivorous fishes. Journal of Fish Biology. 53:222-35.
Afeworki Y, Zekeria ZA, Videler JJ, Bruggemann JH. 2013 Food intake by the parrotfish Scarus ferrugineus varies seasonally and is determined by temperature, size and territoriality. Marine Ecology Progress Series. 489:213-24.
Smith TB. 2008 Temperature effects on herbivory for an Indo-Pacific parrotfish in Panamá: implications for coral–algal competition. Coral Reefs. 27(2):397-40
Figure 1 There is lots of information in these panels. You could get rid of vertical and horizontal lines to reduce some of the clutter. I also appreciate that you want to summarise as much information on these figures as possible but it’s very difficult to distinguish between some of the species on the left hand panels. Could another colour scheme or different shapes/patterns help distinguish between species?
L312 Caption for Figure 3. I was a bit confused by what are averages and what are medians. Perhaps clarify "bars in panel E and F are average values from X surveys?
Figure 4 As in figure 1 its difficult to discriminate between some of the species here, especially S. rubroviolaceus and Ch sordidus. Maybe patterns or different colours could help
L134-136 are the correlations between biomass and grazing/bioerosion for a linear relationship? If not is it appropriate to make statements about increases in functional rates per kg increase in biomass?
Author Response
The study illustrates the value of having species level information when calculating rates of grazing and bioerosion and provides species specific information on feeding rates that will improve estimates of these processes in the future. Overall, I think it’s an excellent addition to the special edition that will provide valuable information for understanding herbivory, bioerosion and carbonate budgets on coral reefs. I have only a few comments for the authors to consider below.
R: Thank you for your thorough and positive review.
L112 feeding rates will change throughout the day. I assume surveys were evenly spread across this time period with no bias towards surveying a particular size or species? Is it possible to illustrate this in supplementary files?
R: Yes, feeding rates were quantified haphazardly, mainly in the times 10-12 and 13-15 to avoid lower rates at the beginning and end of each day as well as potential peaks during midday. Time of each observation is visible in the raw data tables which will be available over a link (but which unfortunately could not be uploaded to the Diversity review forum). As we have mostly recorded time slots we have not plotted or tested the effect of daytime on bite rates in this study.
L134-135 The assumption here is that there were no scars in the survey area before the feeding bout?
R: Scars were only measured if they could be attributed to the observed fish, based on the occurrence of fresh scars and the number of bites observed. Fresh scars are very white and thereby distinct from older scars (which are already overgrowing with CCA or turf algae after few hours-days).
L166 As study locations are close to the equator I assume season and variation in water temperature are not major issue here when calculating grazing/bioerosion rates. But seasonal and spatial differences in temperature could influence feeding rates (see refs below). How would these factors be considered for calculating grazing/bioerosion at locations at higher latitude?
R: Yes, due to low variability in daylight (30 min) and temperature (~2°C) at our study sites we assume the seasonal influences on feeding rates are minimal. However, this is an interesting point for other study sites and following your advice we have added references on seasonal variability to the introduction (L. 90) and to the discussion (L. 420). In general, we would advise to quantify feeding rates for the most common species at each study location, in higher latitudes additional seasonal observations would be valuable or a conversion factor could be derived from the available literature.
Figure 1 There is lots of information in these panels. You could get rid of vertical and horizontal lines to reduce some of the clutter. I also appreciate that you want to summarise as much information on these figures as possible but it’s very difficult to distinguish between some of the species on the left hand panels. Could another colour scheme or different shapes/patterns help distinguish between species?
R: It is true that these panels depict a lot of information but in our opinion it is very valuable to plot all data in one plot to directly compare species rates. Following your suggestions we have deleted the grid lines and further optimized the colour scheme to better depict differences between species for Figures 1, 2 and 4. We also provide tables with size-specific feeding metrics and functions in Table S1 and Table 2 which are valuable for species comparison.
L312 Caption for Figure 3. I was a bit confused by what are averages and what are medians. Perhaps clarify "bars in panel E and F are average values from X surveys?
R: We have followed your suggestion and changed the figure caption to “Bars in E-F depict average values. Boxes in A-D [...].”
Figure 4 As in figure 1 its difficult to discriminate between some of the species here, especially S. rubroviolaceus and Ch sordidus. Maybe patterns or different colours could help
R: Following your above comment we have optimized the colour scheme to emphasize differences between S. rubroviolaceus and Ch. sordidus. Also, these species are different in size and lines can be differentiated on that basis.
L134-136 are the correlations between biomass and grazing/bioerosion for a linear relationship? If not is it appropriate to make statements about increases in functional rates per kg increase in biomass?
R: Yes, the correlations are linear and tested with Pearson’s product-moment correlations (as stated in L. 222 and see correlation formulas in the panels). The reason the lines in Fig. 5 are curved is because they are plotted on a logarithmic scale (as stated in the caption).
Reviewer 2 Report
The present study have shown the species-specific difference in parrotfish grazing ability and their bioerosion functions in the Indian Ocean. Clear size-specific and species-specific variations in grazing ability and bioerosion functions are shown.
I have some minor concerns as follows:
1) L16:
“…new settlement space through grazing….” should be revised as “new settlement space for benthic animals through grazing….”
2) L116-L117:
The authors stated as “Studied species were the most common parrotfishes found in central Indian Ocean reefs”. If the authors have some references, they should cite them here. Or, if the authors have any concrete data about the parrotfish density in the study sites, the data should be shown as a supplementary file.
3) Figure 1A, 1C, 1E, 1G:
Since some colors of dots were quite similar among species (e.g. Cetoscarus ocellatus vs. Ch. stronglyocephalus), it is somewhat hard to detect the species-specific differences. The authors should use distinct colors or symbols (e.g. circle, triangle and square) among the nine species.
4) Figure 1B, 1D, 1F, 1H:
Since the background is white, yellow dots and letters are hard to read. Please use another color (e.g. red).
5) L297-298:
In these lines, the authors used “Maldives” and “Chagos Archipelago” to describe the study sites. In contrast, the authors used “Gaafu”, “Peros Bathos” and “Salomon” to describe the study sites in figures and other sentences. Since readers are not necessarily familiar with the name of the study sites, it does not seem that using the local names (Gaafu, Peros Bathos and Salomon) is appropriate. I feel that the authors would like to show the regional difference in parrotfish grazing ability and bio erosion functions. Thus, I recommend that the name of study sites should be shown as follows:
Gaafu => Maldives
Peros Bathos => Chagos Archipelago 1
Salomon= > Chagos Archipelago 2
6) L316-318:
The authors stated the spatial difference in the density of five parrotfish species. However, no data were shown. The authors should show the species- and size-specific difference in the density of nine parrotfish as figure (e.g. bubble plots on Figure 1C, 1D and 1E).
7) L322:
Since readers are not necessarily familiar with the study sites, it does not seem that using the local names (“Kandahalagalaa”) is appropriate. I recommend that the study sites should be shown as follows:
Kandahalagalaa => one study site in Maldives
8) Figure 4:
Some colors were quite similar among species (e.g. Cetoscarus ocellatus vs. Ch. stronglyocephalus). Please refer comment 3.
9) Figure 5:
For study site name, please refer comment 5.
Again, the color combinations should be changed (e.g., blue, red and green).
10) L449-451:
Although the authors stated the spatial difference in the density of four parrotfish species, no data were shown in the manuscript. Please refer comment 6.
11) L460, 462, 463:
“m2 m-2 yr-1” should be revised as “kg m-2 yr-1” ?
12) Figure S1-B:
What is “Male” and black dot?
13) Caption of Figure S2:
“>” should be revised as “<“ or “p<“.
Author Response
The present study have shown the species-specific difference in parrotfish grazing ability and their bioerosion functions in the Indian Ocean. Clear size-specific and species-specific variations in grazing ability and bioerosion functions are shown.
R: Thank you for your thorough and positive review.
I have some minor concerns as follows:
1) L16:
“…new settlement space through grazing….” should be revised as “new settlement space for benthic animals through grazing….”
R: Thank you for this comment. We feel the addition would be distracting and not necessary in the abstract but we have added the suggested information to L. 48 “[...] provision of settlement space for corals and other benthic organisms by clearing benthic substrate [...]”
2) L116-L117:
The authors stated as “Studied species were the most common parrotfishes found in central Indian Ocean reefs”. If the authors have some references, they should cite them here. Or, if the authors have any concrete data about the parrotfish density in the study sites, the data should be shown as a supplementary file.
R: We were meaning to say that we studied the species that we most commonly observed. We have changed the sentence accordingly to read “Studied species were the most common parrotfishes on the visited central Indian Ocean reefs”. Data on parrotfish abundance at each site can also be found in the raw data tables that will be published with the article (but which unfortunately could not be uploaded in the review forum in Diversity) and in the references cited in L. 185.
3) Figure 1A, 1C, 1E, 1G:
Since some colors of dots were quite similar among species (e.g. Cetoscarus ocellatus vs. Ch. stronglyocephalus), it is somewhat hard to detect the species-specific differences. The authors should use distinct colors or symbols (e.g. circle, triangle and square) among the nine species.
R: We have optimized the colour schemes in Fig. 1, 2 and 4 to better reflect the differences between species. Additionally, we provide species and size comparisons in Tables S1 and Table 2.
4) Figure 1B, 1D, 1F, 1H:
Since the background is white, yellow dots and letters are hard to read. Please use another color (e.g. red).
R: Yellow depicts scrapers throughout the manuscript (e.g. Fig. 4, where blue+yellow=green) which is why we do not want to change the colour to red. However, following your suggestion we have changed the yellow letters to a darker shade to increase readability. Also, we have deleted the grid so the letters and dots are easier to read.
5) L297-298:
In these lines, the authors used “Maldives” and “Chagos Archipelago” to describe the study sites. In contrast, the authors used “Gaafu”, “Peros Bathos” and “Salomon” to describe the study sites in figures and other sentences. Since readers are not necessarily familiar with the name of the study sites, it does not seem that using the local names (Gaafu, Peros Bathos and Salomon) is appropriate. I feel that the authors would like to show the regional difference in parrotfish grazing ability and bio erosion functions. Thus, I recommend that the name of study sites should be shown as follows:
Gaafu => Maldives
Peros Bathos => Chagos Archipelago 1
Salomon= > Chagos Archipelago 2
R: You are correct, we would like to show both broad differences between countries and regional differences within the Chagos Archipelago and are therefore comparing atolls. As we have described the study sites in the method section (L. 179: “at Gaafu Dhaalu atoll, Maldives (n=5) [...] and at ten fore reef sites in the Chagos Archipelago across the atolls of Peros Banhos (n=6) and Salomon (n=4)”), and again in the results section (L. 288: “Average ecoystem functions were [...] compared across the atolls of Gaafu Dhaalu (Maldives), Peros Banhos and Salomon (both Chagos Archipelago)”), and again in the Figure 3 legend (L. 310) we feel that using the actual names of the atolls is more accurate than using Chagos Archipelago 1 and 2. Also, the map of all study sites clearly shows reef sites, atolls and locations (Fig. S1).
6) L316-318:
The authors stated the spatial difference in the density of five parrotfish species. However, no data were shown. The authors should show the species- and size-specific difference in the density of nine parrotfish as figure (e.g. bubble plots on Figure 1C, 1D and 1E).
R: Total parrotfish densities at each atoll (also size-specific) are shown in Fig. 3 C and E. We have thought about bubble plots depicting species- or functional group-specific abundances in the first version of this manuscript, but as we display many Figures already we have decided to instead describe this result in the text. However, following your suggestion we have now added density data for the most abundant species in the text (L. 317-321). Other data on species-specific abundances is available from the raw data tables which will be available with the manuscript.
7) L322:
Since readers are not necessarily familiar with the study sites, it does not seem that using the local names (“Kandahalagalaa”) is appropriate. I recommend that the study sites should be shown as follows:
Kandahalagalaa => one study site in Maldives
R: We agree and have removed the site name as suggested.
8) Figure 4:
Some colors were quite similar among species (e.g. Cetoscarus ocellatus vs. Ch. stronglyocephalus). Please refer comment 3.
R: Following previous comments we have updated the colour scheme on parrotfish species for all Figures to increase differences between species.
9) Figure 5:
For study site name, please refer comment 5.
Again, the color combinations should be changed (e.g., blue, red and green).
R: We appreciate the suggestions on names and colours, but for the reasons we described above we decided to stick with the used combinations.
10) L449-451:
Although the authors stated the spatial difference in the density of four parrotfish species, no data were shown in the manuscript. Please refer comment 6.
R: The data to support this statement (“key species for grazing function are Ch. sordidus and S. niger due to their high abundances, and Ch. strongylocephalus and C. ocellatus due to their large body size”) is shown in Figure 4 (=contribution to total grazing and erosion functions). We have now also added densities of species in L. 317-321 (see comment 6). All data is available from the raw data table.
11) L460, 462, 463:
“m2 m-2 yr-1” should be revised as “kg m-2 yr-1” ?
R: The units in these statements are correct as they state grazing estimates (m2 m-2 yr-1) and erosion estimates (kg m-2 yr-1).
12) Figure S1-B:
What is “Male” and black dot?
R: Malé is the capital of the Maldives and serves as a reference point.
13) Caption of Figure S2:
“>” should be revised as “<“ or “p<“.
R: Thank you, we have changed this to ” ‘***’ <0.001, ‘**’ <0.01, ‘*’ <0.05, ‘ns’ not significant.”
