Insect Abundance and Diversity Respond Favorably to Vegetation Communities on Interim Reclamation Sites in a Semi-Arid Natural Gas Field

Round 1

Reviewer 1 Report

I have no more questions. All requests were met. Please change Artemisia tridentata ssp. Wyomingensis with Artemisia tridentata ssp. wyomingensis.

Author Response

Thank you for your time.

We have made your suggested change to keep wyomingensis lowercase. 

Reviewer 2 Report

Comments for author File: Comments.pdf

Author Response

Thank you for your time and review.

We have added in the caption of Figure 1 that a map of the Pinedale Anticline gas field can be downloaded from Pinedale Anticline Data Management System.  We understand your concern, though to make a Figure with the spatial extent of the gas field, we would need at minimum 1 full page and we believe it would detract from the results of the study. The purpose of the Figure is simply to show readers where the study area is located. 

 

We have made all other changes that you have suggested and appreciate them.

Reviewer 3 Report

Dear Authors

The study is interesting but needs many questions to answer before consideration. Introduction, M&M, Results, Discussion and Conclusion are still need attention. Particularly, authors should work on the M&M and results section.

Comments are attached here with attachment. Please find it for your concern.

Comments for author File: Comments.pdf

Author Response

We have added several more current references to the introduction section, though several of the 'older' literature we had initially cited are seminal papers and we did not remove those references.

Your point is well taken about our sampling window being small.  We addressed this with a reviewer in our last revision.  We suggest in the discussion that longer sampling would be beneficial.  That said, the take home message from our data over 2 years (with 16 sites per year) was similar with 7 of the 8 hypotheses we tested being backed by p values <0.05 and the 8th with a p = 0.07.

We added the National Audubon Society guide used to assist Dr. Smith in IDing insects.

We have addressed your other comments in the M&M and Results.

We were a bit confused regarding your comments about the discussion, but we made some changes there and hopefully addressed your concerns.

 

Thanks for your time and your review.

Round 2

Reviewer 3 Report

Dear Authors

Thank you for your changes in the manuscript. My queries concerning the MM and results are still same as in 1st version.

According to your answer "That said, the take home message from our data over 2 years" What is meant by this?

We added the National Audubon Society guide used to assist Dr. Smith in IDing insects." I asked about the keys to be used in the identifications?? Can you provide some insects images as supplementary material??? 

Please answer to every query in the report, do not make it general. 

Article needs a lot of changes in the current version before consideration. 

In the Table 1 and 2, How did you make the possibility of using ANOVA along with percentage??? I am literally strange in this about it. 

Author Response

We apologize, as we attempted to answer all of your questions last round, though after hitting submit button on the 'reply to review report' we could not go back and edit it and therefore many of our comments were in our note to the editor.

What we meant by our statement 'take home message is the same over 2 years' is that in our 2 year study, we found significantly higher abundance on all of our reclaimed well pads compared to adjacent reference areas.  Additionally, all reclaimed sites with Rocky Mountain bee plant had higher insect abundance than sites with grass.  In our richness tests, 3 of 4 tests (2 each year) had highly significant differences with insect family richness being higher on reclaimed sites than reference sites -- in the one that wasn't highly significantly different, our p-value was 0.07, so there is an argument (p<0.10) that there still was a difference.  Additionally, our abundance numbers are so much higher (several folds) on reclaimed sites than reference areas in both years of the study.  Finally, we address this in the discussion -- it is common for mass flowering literature to be limited to short sampling windows within a growing season... additionally, Longcore (which we cite) is a seminal paper regarding insects as indicators of restoration success and he (along with numerous other papers) highlight a key advantage of studying insects is that a large (statistically valid number) can be collected in relatively short periods of time compared to other wildlife.

Dr. Smith used a dichotomous key with the addition of help from the Audubon book.  We added that he used a dichotomous key.  I have searched through numerous manuscripts related to insects and do not see any of them (especially not in journals specifically intended for entomologists) with supplemental photos of images.  

We do not see any new report here.  When you say 'do not make it general', we are confused as many of the statements in your report are general.

We did not use an ANOVA (perhaps this is why you are confused?).  We used a chi-square test.  I have changed the wording in the captions of Table 1 and 2 to clearly illustrate that we converted number of pixels (out of 2880 to percentages), and I bolded/underlined the cover types which were higher on site types and responsible for differences to further illustrate this. We also added (in our last revision, not this one) that our method was consistent with a previous paper (i.e., Curran et al. 2019 which is the first paper to utilize a chi square analysis on reclamation/reference sites utilizing samplepoint - our statistical methods here are identical to that paper which has been cited 8x in Restoration Ecology).

Round 3

Reviewer 3 Report

Dear Authors

I suggest again to review the Line 181-221 for changes as per recommendations. Because after publication it becomes the property of the authors and vagueness will hurt authors dignity.

Author Response

Since there is no attachment here, we are assuming the recommendations you refer to are from the .pdf attached to your first comments in this round of revision.

Specific to lines 181-221... we have addressed these all in our last two responses.

Your comment that we had a short sampling frame is well-taken.  We have already stated that as a limitation in our discussion, and this limitation is very common to other mass flowering literature.  Our findings are consistent over the course of two years, with very high confidence (especially related to insect abundance).

We have added one bit above line 181 to clarify that insects were examined 'in a laboratory' at University of Wyoming.  We have stated that an author of this paper, who is a PhD in Entomology, used a dichotomous key and the field guide which he used to help ID insects.  

We better explained our chi square analysis after your first session of comments as requested.

We clearly explain why we grouped insects by vegetation type rather than make a direct comparison (this is in response to your 'why not?' question on line 203 in your .pdf from last revision, which is now line 208.  This is common in insect monitoring, and we had already provided 2 citations to address it.

We adjusted our Tables 1&2 and modified the caption to better inform readers.  We (based on your first review) clarified that this is in accordance with previously peer reviewed literature.

Thank you for your review.  We now feel that our questions about our dignity will not be an issue.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

This paper looks at 16 reclaimed well pad plots that were planted with an annual flower and compares them to 16 adjacent ‘reference plots’.  The reference plots changed between sampling years. Additionally, there were some transects used for vegetation sampling.  The authors however did not attempt to correlate insect abundance and family richness to vegetation cover or richness.  They used categorical variables for the sites. 

As written, the main take away from this paper is that sites with annual flowers have more floral visitors that sites with native grass. This is absolutely not worth publishing, nor does it have anything to do with restoration or native areas. The authors did collect a lot of data, but the framework of ‘restoration of well pads’ increase insects is ridiculous.  The authors start the monumental overselling of their data by referring to the UN and the decade of ecosystem restoration and natural resource extraction in western NA.  They look at one site planted with 1 annual flower (for reference, there are thousands of native flowers to the northern Great Plains of North America- most of which are not annuals).  Ecological restoration aims to assist the recovery of an ecosystem that has been degraded and Restoration ecology is the scientific study of repairing disturbed ecosystems.  Planting a single annual flower in no way has any basis in restoration ecology and is not ecological restoration.  With their data the authors demonstrate that the ‘restored’ sites are not close in composition or structure to the reference community.   

If there is some actual ecological hypothesis that these data can be used to test, then the authors might be able to write a publishable paper.  Right now is it as novel and impactful as saying my flower bed has more floral visitors than my lawn- astonishing!

Reviewer 2 Report

The review of the manuscript number 906773, entitled: "Vegetation Communities on Reclaimed Well Pads Increase Insect Abundance and Diversity in a Semi-Arid Natural Gas Field".

General comments

The manuscript is concerned with the analyse insect responses to reclamation efforts in the Pinedale Anticline natural gas field in  semi-arid Wyoming. In my opinion the discussed issue is very interesting, especially because of the strong environmental degradation and visible loss of biodiversity of such regions. There is a strong need for research on effective land reclamation and ecological restoration methods that give the opportunity to restore natural biodiversity and stop habitat degradation. However there are some obscurity in the manuscript. My recommendation for the current submission is to Minor revision. A list of comments with line numbers is given below.

Title

The title is clear and properly reflect the content.

Abstract and key words

The abstract is suitable and basically good.

Line 18: You use "ecological reclamation". I think should be "ecological restoration". Please improve it. There is also double space.

Line 19: Once again double space, the same situations are in line 21, 23, 27, 28, 30.

The key words are correct.

Introduction

The introduction is comprehensive, but needs some minor corrections.

Line 59: double space, please improve it and check the whole manuscript.

Line 109-113: Please use the same font size as throughout the text.

Line 114-138: The aim of the study is described too much detail and there are fragments which are mixed with the research methodology. The aim of research and hypotheses should be transparent. I propose to short this paragraph and simplify the aims.

Line 122-123: Which previous studies? Maybe you should add a references...

Line 124-127: You write " Based on previous ‘mass flowering’ literature, coupled with knowledge of vegetation structure and color influencing flying insects (see above)...". Which literature? When you use sentence like this I propose to add some the most important references conforming the sentence.

Methods

What was the distance between the particular studied sites. You don't mention it.

Please add the information about the statistical software which you use for data analysis. Did you check the distribution of the variables. You don't mention it.

Results

Line 224-226: I do not see this significance differences, because you do not add a table with any statistics... Please improve it.

Line 226-229: The same as above. I do not see statistics in the table 1 and 2.

Line 229-231: The same comment as above.

Line 235-237: Please use the same font size as throughout the text.

Line 243-255: You have to add the table with statistics. In my opinion the results are not complete at this point .

Figure 3: I propose to remove the grid lines from the chart. Did you prepare a regression plot for only two point (2015, 2016)?

Line 260-272: The same comment, for me results without statistic values (e.g. F or another and p values) are not complete.

Figure 4: I propose to remove the grid lines from the chart.

Line 302-305: "significant distribution" ? I do not see this results...Please improve it.

Discussion

Line 321: Double space, please improve it.

Line 325: Double space, please improve it.

Line 372: Double space, please improve it.

Reviewer 3 Report

In this manuscript, the authors have measured arthropod abundance and family richness, together with several vegetation parameters, in reclaimed well pads characterised by mass flowering Rocky Mountain bee plant and adjacent reference habitats characterised by native grasses. Although the focus of this work is interesting, several major issues make this manuscript unsuitable in its current form, but that if addressed could greatly improve its value.

 

The rationale for this study needs to be better presented.

In the introduction, the authors refer to fragmentation and ecosystem services, but both concepts are loosely related to this work. The habitats investigated here it’s not just a fragmented habitat, and while the authors obtained biodiversity measures (family richness), they did not quantify ecosystem services. These detours from the focus of the study make the introduction confused and claims in the discussion about ecosystem services pointless. My suggestion is to start directly presenting these disturbed habitats and on the possibility of supporting the arthropod community by restoring the original habitats.

 

The functional analysis is based on a single trap** technique that does not work equally well for all arthropod groups and hence, the results are biased.

To collect arthropods, the authors used sweep netting, a technique which is suitable for pollinators, herbivores, and few predators such as robber flies. Such a trap** system is not really suitable for most decomposers and ground-active predators, for example. Therefore, the results of the functional analysis are biased by the trap** system adopted, and cannot reflect the whole arthropod community. My suggestion is to remove entirely this part.

 

Some important details about the experimental design are missing.

It is essential to mention how distant were the sampling sites from each other to understand if they were far enough to guarantee independence. Many pollinators could fly over relatively long distances. Moreover, the sampling effort was not that great (just 4 half days over 2 years), and concentrated in September. In the introduction, the authors should justify why they have only sampled at the end of the season and also on which groups were they really focusing (I’d say bees and other pollinators…). There has to be an ecological justification for such a short sampling window.

 

The taxonomical resolution is very low.

The authors have identified the material only until the taxonomical level of family. This is unfortunate because it seriously limits the importance of this work to understand better arthropod communities in this system. Would they be able to provide a more informative description of the arthropod community if they were given more time?

Another disturbing note is that the authors always refer to “insects” while presenting their data, which include also spiders. Please, change “insects” to “arthropods” throughout the text.

 

The statistical analysis is not sufficiently explained.

I have some reservations about the linear mixed models for insect abundance and richness. Given that the sampling effort was low, I would not be surprised that those responses were not normally distributed (which is common for arthropod populations). Did the authors checked for this and other assumptions, and eventually how? Did they make any model selection and validation? None of this is mentioned in the text. I would like to see the model residuals to have a better idea of the reliability of those models. Additionally, the authors must indicate which software was used for the statistical analysis.

Finally, it is unclear why for certain analyses it was necessary to include a random factor (which is not clearly mentioned, but I suspect it to be “site”), while for others (the chi-square test), this was not needed. Why not using linear mixed models for everything? As I said, I struggle to believe that those data were normally distributed. The authors should also clearly specify in the statistical analysis if they focused only on the most represented groups in the analysis, and not in the results.

 

The figures have to be improved.

Stacked bar plots are known to be a poor way to represent this kind of data (see Edward Tufte’s extensive work on visual communication), and many small values are impossible to be read from the graphs. Dot charts would have been much more informative. Also, alphabetical order does not make much sense here, and either a grou** by taxonomical order or by the functional group would have been better.

It does not make much sense to consider Year as a continuous variable when you only have two sampling years. A regression line in Figs.3 and 4 is misleading. Change it to points (estimated mean ± 95% CI). Also, “arthropod abundance” would be a better label on the vertical axis in Fig. 3, and “family richness” a better label for the vertical axis in Fig.4, as there can be many types of richness.

As mentioned above, I don’t trust the results on functional traits and I would exclude this part from the ms. Moreover, I don’t see the reason to present these data in two separate figures (Figs.7 and 8). Also, “Ecological_Role” is a bad label. Please change it to “Ecological role”, or just leave it blank.

 

Minor comments:

The keywords are mostly OK, but "ecological restoration" and "ecosystem restoration" are too similar, just pick one of the two.

The title mentions “insect" but in this study the authors also considered spiders. In some of the figure captions, the authors mentioned insect abundance, but spiders are not insects…

The abstract requires major revision, as in its current form the rationale of the study is not clearly identified, it is lacking of quantitative results (almost no results in general), the hypotheses of this work were not stated, and the conclusion from this work are missing.

Reference list

Please, provide details for ref. #37.

In ref. #50, the Latin names should be in italics.

In general, the format is mixed (e.g., sometimes the issue number appears followed by a colon).

The table in the supplementary material does not have a title, which is not acceptable. Also, the table header should be bold and the table families rearranged according to the taxonomical order. The alphabetical order is rarely a good option. There is also a typo in “detritivore”.

 

Please, find additional comments in the pdf of the ms.

Comments for author File: Comments.pdf

Reviewer 4 Report

This study, concerning the monitoring restoration of semi-arid natural gas, regards an interesting topic but it needs many improvements. Starting from the introduction there are many papers that report results about restoration of seminatural vegetation considering insects L.54-55 should includes more citations. L 108- 113 the study does not deals with the effect of grazing since all the study site are grazed (I presume) so this part should be changed. Methods should be improved with the addition of details on type of management receive from reclaimed areas: seed addition? Natural succession? Details about grazing are lack. Fig. 1 is not informative at all and disproportionate. Statistical analysis needs clarifications: why mixed effect model? Which were/was the random factors/factor? Why the complete results of analysis were not reported in Result section?  Discussion should be improved, overall, from table 1 and 2 I see that bee plant reference and grass reference had quite different characteristics regarding bare ground cover, herbaceous cover and rock percentage. So, I think they were different grassland is not possible, in my opinion to separate the effect, on insect abundance, due to the natural differences between the two communities from that of the closeness to the beeplant reclaimed sites. The introduction of a covariable such as rockiness or grass cover could help to clarify this aspect.  The reclaimed beeplant sits also had a high percentage cover of non-native plant, this aspect should should be deepened in discussions to understand which type of reclamation is better in a broader ecological meaning than just considering insects presence.