The Occurrence of the American Burying Beetle (Nicrophorus americanus) and Associated Silphid Beetle Community in South Dakota: Implications for Managed Relocation

Round 1

Reviewer 1 Report

Results and discussion should be carefully separated; sub-sections would better structure composition.  Habitat ranking seems arbitrary. Authors use inappropriate population estimate method and fail to discuss the weaknesses of their approach. All parts are written without much flow, have redundancies, and lack transitions. Discussion repeats too much from the introduction and provides new info that belongs into intro. Trap design and marking techniques should be illustrated with pictures or drawings. Figure S4 belongs into the paper. More details in attached PDF.

Comments for author File: Comments.pdf

Many grammar and style issues. See attached PDF for examples.

Author Response

We appreciate the editing suggestions and incorporated with tracked changes except where noted below in blue.

We prefer the and, style and left it throughout.

In South Dakota specimens (indicating pinned individuals) was left.

The reviewer makes suggestions and seeks clarification about the trap design. We added reference to the original paper that includes illustrations.

The reviewer also asked for clarification of the habitat rating and we added:

Although N. americanus is classified as a habitat generalist, replacement of grasslands with row crop agriculture and arid environments do not support its occurrence (21). Extensive previous sampling in northern Nebraska and South Dakota has led to development of criteria of likelihood to capture N. americanus based on visual characteristics of habitat. To ensure maximum likelihood of capturing N. americanus when present, we selected trap locations based on the following criteria while driving on public roads.

For statistics, we present the Standard Error which is common when presenting ecological data and consistent with previous work on N. americanus.

Page 12, based on Reviewer 1’s comments we added information to the introduction. The paragraph beginning with Silphids constitute… remains important to discussion that follows.

Page 14, carrion beetles use maggots from flies and as such do not provision for their larvae.

We added details on the number of beetles per acre and the classification of “Prime”.

We appreciate the careful editing to improve readability and clarity.

Reviewer 2 Report

Review of diversity-2908716.

Dear Editor,

The manuscript deals with the distribution of an endangered insect species Nicrophorus americanus. The study is up-to-date and relevant for publishing in the journal Diversity.

Comments and questions to the authors:

There are no line numbers, I don’t know if this is an issue because of the editor or the authors.

Introduction: a picture of the species and its habitat could be added?

p.2: do not use “N.americanus” at the beginning of a sentence but list the full name Nicrophorus

Fig.1: why was Tripp county the “best for N.americanus? Can you provide a habitat map or some freely available satellite or aerial image why is that? Or is it because you sampled mostly this county?

see the attached PDF for additional comments

Comments for author File: Comments.pdf

Author Response

We greatly appreciate the comments and suggestions made by both reviewers and believe that our manuscript is much improved.  We track changes of the two reviewers and detail how we addressed suggestions below.

Reviewer 1.

We tracked changes in red and made all suggested edits. 

We added to the first part of the sentence to include potential benefits to warmer temperatures.

We chose not to include yearly data because our sampling occurred in different areas, producing few N. americanus in 2019.  We added the following sentence to explain our decision to use percentages. 

Because sampling in 2018 and 2019 were to establish South Dakota distribution and 2020 was to estimate populations in the best areas, we summarized captures by percentage across years.

We updated Figure 1 to make it legible (this was an image resolution issue)

Similar to the previous response, our sampling design did not allow us to compare yearly data because trap purposes and locations were different among years.  To clarify our approach, we added “Because trap locations varied among years”, a series of exploratory analyses were used to determine the relationship between N. americanus and other silphid beetle species.

The reviewer suggests adding a Table with weather data during sampling.  This would normally be an excellent addition.  However, our sampling occurred over 2 weeks in June and 2 weeks in August each year for three years and covered an area of more that 100 miles (220 km) east-west and 60 miles (130 km) north-south.  We do not believe the weather data during sampling help interpret the results of this study.  Data are available from weatherunderground.com as are state averages for the periods of sampling.    

The reviewer suggested adding life history details to the introduction and we added the following paragraph:

Because of intense inter- and intra-specific competition, carrion beetles exhibit niche partitioning with species-specific seasonal and daily activity patterns (10,11). Nicrophorus americanus is nocturnal, with activity occurring with nighttime temperatures above 13^o C. It has a one-year life cycle with northern populations emerging in late May, breeding in June, and offspring emerging in August (13).  

The reviewer asks about font size which varied after type-setting.  We formatted entire manuscript to 10 pt font. 

The reviewer highlights lack of predators in areas where reintroduction is favored.  We added “and predators”  

Reviewer 1 also asked a number of statistic questions.  We clarify our approach and the basis for our choices here.  We did not add to the manuscript itself.

why have you used AICc, the corrected version of AIC for small sample sizes?

Correct. Burnham and Anderson (2004) recommend the use of AICc when the ratio of data points (n) / the number of parameters (k) is < 40. Here, our ratio is ~16. Moreover, it is suggested that, since AICc approximates AIC at large sample sizes regardless, that AICc should be used as a default. Researchers often conclude that AIC overfits because it fails to use the second-order criterion (see, Burnham and Anderson 2004).

Burnham, K.P. and Anderson, D.R., 2004. Multimodel inference: understanding AIC and BIC in model selection. Sociological Methods & Research, 33(2):261-304.

what about the results of glmm models? You are presenting only the model AICc numbers, but not the coefficients of particular variables. Show at least the best model.

The fixed effects are co-occurrence (abundance) of other silphid species, with N. americanus abundance as the response variable. The top four models are listed with the appropriate model parameters (log likelihood, AIC, delta-AICc, and AICwt) in Table 3 for relative comparison and assessment of the global model. Of note, the GLMM output for the top model is referenced in the paragraph above Table 3, with the parameters of interest, i.e., adj. R^2, beta value, standard error, and p-value.

“The top model for N. americanus abundance included O. inaequale and N. marginatus, which explained 99% (adj. R2) of the overall variation in the silphid species used in the global model. The abundance of O. inaequale displayed a positive relationship with greater N. americanus abundance (β ± S.E. = 0.032 ± 0.017, P = 0.06), whereas N. marginatusabundance had a significant negative influence on N. americanus abundance (β ± S.E. = -0.008 ± 0.001, P < 0.001).”

This could tell what is the probablity of the given model being the best of all the models presented. So it is only 20% chance that this model is the best one? This is quite low, maybe it is because you used AICc instead of AIC?

AICc weight is the relative proportion of the total amount of predictive power provided by the set of models contained in the full model being assessed. In the context of the presented analysis and use of field-collected insect data, I would argue that this is quite high. We’re explicitly testing silphid beetle community abundance to explain the variation in N. americanus abundance, a very narrow approach to explain this variation but appropriate to our objective. We would expect that other fixed effects, i.e., additional abiotic factors, would explain more variation and certainly improve the AICwt relative to the model set considered here. 

Again, we sincerely appreciate the thoroughness of the review and the suggestions. 

Author Response File: Author Response.docx