Riparian Vegetation Structure Influences Terrestrial Invertebrate Communities in an Agricultural Landscape

Round 1

Reviewer 1 Report

The role of forested riparian buffers in structuring the riparian invertebrate communities in a catchment impacted by agriculture

Review:

This manuscript brings engaging research on riparian community. Overall impression of the manuscript is highly positive – interesting topic overlapping to practical application (landscape, river system and agriculture management, general nature conservation), well-structured manuscript, clearly presented results, inspiring discussion and good English. However, I found one methodological problem with functional feeding groups (FFGs) of the invertebrates:

When splitting specimens to FFGs, authors rely on literature listed in Suppl. 1. However, authors sometimes pick up just some findings of these resources and do not follow feeding habits valid for majority of the given taxa, which is essential when determining specimens to the family level (mainly). Actually the family level would usually be sufficient for most taxa – e.g. Coleoptera in this manuscript. But some specimens were placed just in the (sub)order, which doesn’t allow appropriate placement into particular FFG. The biggest problem with Heteroptera. All heteropterans were listed as omnivores, but their diet is usually actually quite narrow – either zoophagous (e.g. Nabidae, Reduviidae, etc.) or phytophagous (Pentatomidae, Coreidae,…) with some families containing both categories (e.g. Miridae – this probably led authors for selecting omnivory for all Heteroptera). Given that Heteroptera were actually very abundant in the presented research (second in FBF, third in UBF- Fig. S4), their appropriate categorization could significantly affect results of FFGs.

Next to Heteroptera, similar problem is with Culicidae and their classification as zoophagous. Actually males are nectar-feeding, females sometimes too and mosquitoes can be important pollinators (see e.g. 10.1073/pnas.1910589117; 10.1146/annurev.en.40.010195.002303). Again, Culicidae were very abundant (fig. S4) and their placement into zoophagous insects could affect the results with FFGs. Another problematical FFG categorization is with Neuroptera and Gastropoda, but these two groups were not so abundant.

 

Thus, my suggestion is to determine the sampled specimens to the family level (Heteroptera, Neuroptera, Gastropoda) and place these groups to the appropriate FFG. Problem should be the mosquitoes. I suggest their placement into herbivore group, as the blood-sucking females possibly don’t affect ecosystem – but this is just my opinion which can be wrong). Generally, the FFGs system could be reconsidered – maybe more detailed groups (with nectar-feeders, blood-suckers, seed-feeders, etc.) or with also considering feeding habits of the larvae (see e.g. Neuroptera, Culicidae). Another option – allowing practical implication for agricultural management – would be just looking for known agricultural pests, their predators and pollinators, and comparison of their abundance in FBF and UBF.

If it will not be possible to redetermine the sampled specimens into the family level, I think the whole FFGs analysis has to be omitted or highly modified – analyzing just taxa with surely known major feeding habits (spiders, most Coleoptera families atc.).

 

Beside this, I found just some minor mistakes – citation 13 (l. 584-586), typing error in l. 237.

Generally, I really like this manuscript, but the FFGs issue could really affect the results and it has to be corrected.

 

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

This study assessed riparian invertebrate community' responses to changes in riparian habitats in an agricultural catchment in Romania. The survey was well conducted and the data would be very useful. The statistical analyses are sufficient but the statistical results are not completed and limited. Therefore, I recommend to minor revisions for the publication.     The statistical results are not completed, such as F-values for PERMANOVA and ANOVA.      Fig.1 should be include the latitude and longitude.    The font size of the figures are very small, especially Fig. 2 and 3. Fig.2 UBF and NMDS1 and other characters were very small size.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 3 Report

This paper has very interesting results that justify its publication; specifically, it highlighted the importance of local riparian buffers in mitigating the effects of land-use on riparian invertebrates’ diversity and their potential to be used as a multi-functional management tool in agricultural landscapes.  The manuscript is written well and described comprehensively.  While I am generally positive about the manuscript, I think there are some suggestion and few minor comments which I hope it will help to improve the manuscript.

 

Major concern

The authors applied various statistical approach such as ANOVA, PERMANOVA, RDA etc. to reveal the forested riparian buffers support a distinct invertebrate assemblage with a clear tendency for higher overall invertebrate diversity.  However, a question arises about some of statistical approach (i.e. ANOVA) why did you apply ANOVA to compare 2 groups between unbuffered and forested buffered sites without post hoc test?  Please explain it at method part or reconsider using ANOVA.

 

Minor comments

Line 70-71: please add more references.

Line 129: I suggest you use ‘a five-level Likert items’ instead of ‘graded from poor (1) to excellent (5).

Line 151: please provide more detailed information or reference about the program (PAST). As I know it has reference paper.

Line 158: 70% - 80% => 70 – 80%

Line 160-161: ‘All these analyses were performed with the PAST 4.03 software’ => move to end of paragraph.

 

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 4 Report

This manuscript tackles an important conservation issue, that of the role that vegetated riparian strips contribute to providing habitat for beneficial insects compared to unvegetated riverbanks. The authors conclude that vegetation does provide some beneficial ecosystem services in this way.

I have no criticisms of the hypothesis being tested and the field sampling has been rigorously designed with more than adequate replication. A comprehensive range of environmental variables have been chosen and measured as well.  However, the assertion that the whole fauna has been sampled is far from the truth. I am amazed that the collection and identification of the invertebrates, the key part of the project, has been designed and carried out in such a perfunctory way as it is on the invertebrate data that the success of the project hinges. I am therefore suspicious of the conclusions drawn from the results, as the fieldwork is flawed. I give my reasons below:

1.      All invertebrates were collected and identified in the field. There is no information as to how this was possible nor is it reported as to whether the collectors/identifiers had identification manuals or microscopic equipment with them in the field. I have never heard before of any invertebrate project that relied for data collection only on hand and net collections and identification only in the field.  This is not a rigorously scientific way to collect the required data.

2.      We are not told what qualifications the field identifiers had to do this work.

3.      No voucher collection of specimens identified has been kept it seems, so that results cannot be verified and certainly not repeated. It is essential that all scientific fieldwork is repeatable by others. Because of this omission, neither repeatability nor verification have be catered for in this project.

4.      By no means was the whole fauna collected as the field workers only collected what was visible to the naked eye and the sampling was only done once on each site in summer. Soil fauna was omitted as was fauna active at other times of year. So lack of completeness is an issue here.

5.      In spite of much published literature pointing out that, for results to be valid in ecological work of this type, all specimens must be identified to species.  This recommendation has not been followed as here the specimens collected were only identified to family. I find this an astonishing failing, as differences in ecological preferences of genera within families can be significant and even between species in the same genus. As a consequence, I am not surprised that few significant differences were observed between treatments.

6.      It is also well documented in the literature that personal differences of field workers, particularly in ability and thoroughness, can bias the results even up to 100% difference between individuals.  It seems there was no system in place in the fieldwork to control for this -a serious failing.

7.      I am also astonished at the large number of authors listed; I counted 11 from six different countries although the fieldwork was done in only one, Rumania.  There is no explanation at the end of the mss listing the contribution each author made to the text. This must be rectified.  

8.      I cannot understand the term “functional feeding groups”.  They are “feeding groups” not functional groups that must, by definition, include more traits.

9.      Fig 1. Good.  Figure 2, as expected, in fact chosen to be so.  Fig. 3, none of the figures seem to be significant. Table 2, again, little significance except in family differences, which is not surprising since the habitat was more complex on the “buffered” sites. Table 3, not surprising also that these environmental variables were significant as they were chosen to be so.

10.   One important topic not addressed is what effect do the buffered sites have on improving water quality and in reducing bank erosion compared to unbuffered sites.  Why was this not tackled?  This is certainly an “ecosystem service”

11.   There is inconsistency in the terms used to describe the fauna.  Sometimes the common names are used and sometimes the scientific names.  Also, the word Coleoptera seems to be used as if the families are not part of this order, i.e. line 60, very confusing.

12.   It is an usual use of the word “buffer”.  It should be defined.

13.   The title is too long and is misleading.

14.   The abstract begins with a back to front sentence, confusing the reader.

15.   There are too many unnecessary adjectives used such as important, very etc.  I have marked them on the pdf attached.

16.   For some unknown reason the letter s is inserted at the end of all kinds of words that are not possessive nouns.  Again, I have marked them to be omitted on the pdf.

17.   No weather or even climate data is included in the text.  This must be rectified.

18.   The riparian zones are not exactly paired, each sampling sites is a pair with one site as a buffer and one not.

19.   Line 128, visually assessed?  What does this mean?  Clarify.

20.   Table 1. Distance 100m from forest block upstream. English!

21.   I did not assess the validity of the statistical tests, as this is not my speciality.

22.   Line 223.  “soil nitification potential”, curious phrase.  Explain.

23.   The total number of individuals and taxa are low (line 237) and I suspect so low that they are not accurately reflecting the differences between treatment and control.

24.   The authors mention Heteroptera in both buffered and non-buffered sites.  Cicadidae are Hemiptera (line 245, 247) so the way these results are reported are confusing and show no differences between sites. The separation of Hemiptera into Heteroptera and Homoptera is no longer valid and has not been so for many years.

25.   Line 260 and all that paragraph.  The authors are not counting “functional” only feeding groups.

26.   Figure 4 and following paragraph.  I am now totally confused by the introduction of a new variable not mentioned before, pasture as well as natural grassland.  Does this mean native grassland?  By the introduction of the term pasture, does this mean it was grazed?  I thought the ecosystems being compared were treed riverbanks and bare riverbanks.  This conflict must be sorted out.

27.   Again, authors must not quote Cicadidae as distinct from from Heteroptera as the separated of Hemiptera into Heteroptera and Homoptera is no longer supported and has not been for many years.

28.   Fig. 5.  At first glance the taxa so seem to separate out according to habitat but only 2 of the 9 taxa are within the blue circle and 2 of the 8 within the red circle giving rather an inaccurate impression of the distinctness of the separation.

29.   Fig 6, again, inconsistent terminology as common names are used instead of scientific names as in previous Fig.

30.   Line 383.  This is the first time ecological services are mentioned as far as I remember and, as they are not defined, should not be included to assist in justifying the value of buffers.

31.   There are three sections to the discussion.  I do not think the results justify this and it is too long.  In fact, it consists entirely as far as I can see, on repeating the results by family and speculating on the reasons for their occurrence and abundance using rather general information on natural history.  This is not what should be in the discussion.  It should include a comparison of the results here with results from other studies and some recommendations for management giving reasons why. In fact, from these results there seem to be benefits for agriculture possibly from both types of bank vegetation, trees and shrubs versus grass.  However, there will only be benefits if the fauna from the banks move onto agricultural land and this has not been shown.  In fact, I suspect the fauna will not disperse from a habitat that suits them into one that probably doesn’t.  We are left therefore with the only likely benefit from vegetated banks is a possible reduction in soil erosion and an improvement in water quality.  However, unfortunately this was not tested.

32.   Some of the page numbers quoted in the references seem incomplete.  I have not checked the journal requirements for formatting, but I recommend they be checked.

33.   Finally, the English is generally poor throughout and must be rewritten using the services of a native English speaker.

Comments for author File: Comments.pdf

Author Response

Please see the attachment

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The revised manuscript seems much improved. Omitting the FFD helped to straighten the manuscript and resulting outcome stayed the same. I also appreciate improved M&M and rewritten discussion. I think the manuscript can be published in the present form.