A multi-dimensional dataset of Ordovician to Silurian graptolite specimens for virtual examination, global correlation, and shale gas exploration

Xu, Hong-He; Niu, Zhi-Bin; Chen, Yan-Sen; Ma, Xuan; Tong, Xiao-Jing; Sun, Yi-Tong; Dong, Xiao-Yan; Fan, Dan-Ni; Song, Shuang-Shuang; Zhu, Yan-Yan; Yang, Ning; Xia, Qing

doi:https://doi.org/10.5194/essd-15-2213-2023

Articles | Volume 15, issue 5

https://doi.org/10.5194/essd-15-2213-2023

© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/essd-15-2213-2023

© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 15, issue 5

Data description paper

|

01 Jun 2023

Data description paper |

| 01 Jun 2023

A multi-dimensional dataset of Ordovician to Silurian graptolite specimens for virtual examination, global correlation, and shale gas exploration

Hong-He Xu, Zhi-Bin Niu, Yan-Sen Chen, Xuan Ma, Xiao-Jing Tong, Yi-Tong Sun, Xiao-Yan Dong, Dan-Ni Fan, Shuang-Shuang Song, Yan-Yan Zhu, Ning Yang, and Qing Xia

Download

Final revised paper (published on 01 Jun 2023)
Preprint (discussion started on 12 Nov 2021)

Interactive discussion

Status: closed

CC1:
'Comment on essd-2021-280', Zhibin Niu, 14 Nov 2021

Our interactive visual explore of graptolite specimen image data (FSIDVIS) can be accessed through http://fsidvis.fossil-ontology.com:8089/ . Hit the spacebar to view the details of graptolite specimen. We also welcome comments and suggestions of the visual exploration tool.

Citation: https://doi.org/10.5194/essd-2021-280-CC1
- AC1: 'Reply on CC1', Honghe Xu, 18 Nov 2021
  
  Thank you for your comments.
  The fossil specimen image dataset visualization (FSIDVIS) system is for everyone who needs detailed and scientific information of fossil specimens. We will add more data of fossil specimens to the system.
  
  Citation: https://doi.org/10.5194/essd-2021-280-AC1
CC2:
'Comment on essd-2021-280', Zhaohui Pan, 14 Nov 2021

Graptolites are widely used as index fossils for determining rock ages and biostratigraphic correlations from the Middle Ordovician to the Telychian of the Silurian period. 1,550 high-resolution images of 41 genera and 16 families of Graptolites were never seen before. This is a high-quality dataset. Images with high quality in this dataset are well helpful for geological surveys and field practice in China. However, this is a dataset instead of a handbook. There are limits in the fieldwork. Why not develop an automated classification application or an e-book on a cellphone based on these graptolite data? Did you try the automatic classification？

In addition, there is a photo scale in Figure 3. But images in Figure 4 have no plotting scales. Do they have the same scale？

Citation: https://doi.org/10.5194/essd-2021-280-CC2
- AC2:
  'Reply on CC2', Honghe Xu, 18 Nov 2021
  
  Thank you for your comments.
  Graptolites were originated in the Cambrian and extinct in the Devonian. As a kind of extinct organism, graptolites are significant for the role of dating sediments and macro-evolution research. In the study and the dataset we released, we only chose 113 species of graptolites from the mid Ordovician to early Silurian of China. these species of graptolite, though being a small amount of the whole group, are important to shale gas exploration and biozone recognizing. The specimens images of these species have their potential significances to shale gas exploration.
  The uploaded dataset includes high-quality images, and every image shows clear and specific morphology of fossil organism and has its scientific name. We believe they can, to some extent, meet the demands of shale gas bed determination. We added scale bar to some specimens, but not illustrated them in the manuscript. That is why you see rules in the figure 3. We are also training the artificial intelligence (AI) model of automated classifying and identifying, the demonstration version is online (http://aifossil.fossil-ontology.com/#/). To the AI model, there is no need to show the scale bar.
  The fossil specimen image dataset visualization (FSIDVIS) system is for everyone who needs detailed and scientific information of fossil specimens. But now the limitation is that we only uploaded and prepared data of 1550 graptolite speciemsn. Of course this is just a start. More specimens and the related e-book or searching system for mobile devices will follow.
  
  Citation: https://doi.org/10.5194/essd-2021-280-AC2
  - CC3: 'Reply on AC2', Min Zhu, 19 Nov 2021
    
    That sounds great. Please inform me when you publish the e-book or application on mobile devices.
    
    Citation: https://doi.org/10.5194/essd-2021-280-CC3
CC4: 'Comment on essd-2021-280', Ting Nie, 19 Nov 2021

It’s thrill to see the work about the database guarantees multidimensional information of the fossil records. No doubt that the database empowered the research of palaeontology and the application in the shale gas exploration. More are supposed to come to publish in the future. Looking forward to the development of data-driven research in the field of palaeontology.

Citation: https://doi.org/10.5194/essd-2021-280-CC4
RC1:
'Comment on essd-2021-280', Anonymous Referee #1, 23 Nov 2021

The comment was uploaded in the form of a supplement: https://essd.copernicus.org/preprints/essd-2021-280/essd-2021-280-RC1-supplement.pdf

Citation: https://doi.org/10.5194/essd-2021-280-RC1
- AC3: 'Reply on RC1', Honghe Xu, 07 Dec 2021
  
  Thank the reviewer’s comment and the detailed supplementary file. We specifically revised the manuscript and hopefully we will have it submitted soon.
  We re-organized all images and assured that every specimen image shows with the scale in its own photo or separated photo. Now the updated dataset includes 2951 images and the whole size is 10.4 G. every scaled photo file is named with a postfix S. The image name consists of the number of specimen and the species name. one only troubleis that the volume of the whole dataset is quite large. It takes a while to upload and download all files.
  The purposed of our study is not only providing an access of the gallery of graptolite specimens, but also showing that the image dataset is usable in data analysis and visualization and potentially using artificial intelligence.
  The FSIDvis is an interactive visual explorer of graptolite specimen image data (FSIDvis), which is accessed through:
  http://fsidvis.fossil-ontology.com:8089/
  Hit the spacebar to view the details of graptolite specimen.
  TSNE, or t-SNE (t-Distributed Stochastic Neighbor Embedding) is a technique for dimensionality reduction that is particularly well suited for the visualization of high-dimensional datasets that our specimens image dataset belongs to. We tried to analyze these data using this method and show a feasible way to classifying these specimens based on images only.
  In the revised version we will invite a native English editor to proofread the whole manuscript. Thank you for your kind comment.
  Some specimens were collected several years ago. Some of them were renamed and transferred. Species emendations were based on suggestions from experts of Nanjing Institute, we will give their named in the revision. We will make a further revision and state all the related issues.
  
  Citation: https://doi.org/10.5194/essd-2021-280-AC3
RC2:
'Comment on essd-2021-280', Anonymous Referee #2, 04 Feb 2022

The comment was uploaded in the form of a supplement: https://essd.copernicus.org/preprints/essd-2021-280/essd-2021-280-RC2-supplement.pdf

Citation: https://doi.org/10.5194/essd-2021-280-RC2
- AC4: 'Reply on RC2', Honghe Xu, 25 Feb 2022
  
  Thank the reviewer for the comment.
  The main problem comprises two aspects, the dataset quality and the tool. We made the revision accordingly. Furthermore, the whole text is revised in text and writing by colleagues. Hoping this version is much better.
  The dataset is updated and the newly uploaded to share. A new DOI is given in the revision.
  https://doi.org/10.5281/zenodo.619494
  Now it is ensured that every specimen is shown with additional scale bar and that any measurement to the fossil is available.
  The reviewer also stated that some images are not quite clear, or some images seem too empty. These are all problem of image quality. The new dataset, with more hi-resolution images and scale, some microscope images were selected for the case that the fossil itself is small but the whole specimens is quite large.
  The tool, FSIDvis, is updated and more detailed instruction of this software is given.
  The naming system or the method is explained in the text. Every image file is named after its unique number and labeled species name. When the specimens wad re-studied or transferred to another taxon, the file name does not reflect its taxonomic status. a brief revision record is given in the updated tool, FSIDvis. And I also suggest that readers or users of our dataset check our spreadsheet file for detailed information.
  
  Citation: https://doi.org/10.5194/essd-2021-280-AC4

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Honghe Xu on behalf of the Authors (25 Feb 2022) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (01 Mar 2022) by Kirsten Elger

RR by Peter Sadler (16 Mar 2022)

ED: Reconsider after major revisions (26 Mar 2022) by Kirsten Elger

AR by Honghe Xu on behalf of the Authors (29 Apr 2022) Author's response

ED: Referee Nomination & Report Request started (06 Jun 2022) by Kirsten Elger

RR by Peter Sadler (17 Jul 2022)

Suggestions for revision or reasons for rejection

Reviewer’s Background and Summary Review:
I was trained as a taxonomist, but for trilobites and conodonts, not graptolites. I read graptolite systematics extensively as I build and maintain a high-resolution, age-calibrated, global time-line of the species-level macroevolution of the entire graptolite clade. My composite timeline has supported age calibration of the Ordovician and Silurian time scales in addition to several macroevolutionary studies. For the time-line project, I wrote my own data-management software that supports the stratigraphic sequencing program that I co-wrote. Both programs are in Fortran.

Although I rely almost entirely upon co-authors to validate our selection of the graptolite publications that I compile and to vet the quality of the taxonomy, it is likely that my experience is above-average among the potential users that the authors describe for their database. The database should surely be enormously valuable for my time-line projects. I trust this statement clarifies which components of this manuscript that I might be best qualified to review.

This is the fourth version of the image database that I have tested. The authors have made a succession of substantive and successful changes. My previous struggles with the download speed and file format have been essentially alleviated in this latest version. The database is surely ready for a wide user population.

The bulk of this third review focuses on the accompanying manuscript. I will deal with its content. to some extent, but focus primarily on the quality of the English text. The troublesome errors are not substantively improved by the latest edits. Although the quality of English written by non-native English speakers should not reflect upon the quality of the database, I fear that readers might lose confidence in the authors’ attention to detail in the database too. That would be most unfortunate; the database surely deserves better.

In most respects, my prior criticism of the text is still valid. Previously I mentioned the common types of grammatical errors and poor word choices. The one example of subject-verb disagreement that I cited from the introduction was not fixed. The authors’ reply mentions grammar-checking software; it seems to be quite inadequate. This time I have added a marked-up a copy of the PDF file with more edits, corrections, suggestions and questions to guide the authors’ revisions.

In summary, the database files are ready for use by experts, amateurs and students, just as the authors explicitly intend. The manuscript is not yet ready for publication. Obvious inattention to detail in the text file could undermine users’ confidence in the authors’ quality control in the database itself.

Importance of the Image Database
Paleontologic research has been advanced considerably by on-line availability of larger collections of publications than are housed as paper copies in most institutional libraries. This image database is a major step toward an exciting parallel development for paleontological museum collections. The on-line images will not entirely replace expensive travel to examine unique physical specimens or the risky loaning of unique specimens via mail services. They will surely, however, allow more effective preliminary evaluations than the limited photographic plates in printed journals.

Revisions of the Image Database
The large image collection is not set up for on-line browsing. Instead, users download the files to their own computers. Many users will surely welcome this. Three substantive changes to the database structure have made the downloading and searching of the database faster and easier.

1. The image collection has been divided into 49 zipped folders. Individual folders can be downloaded in a fraction of the time that was needed for the single folder; such long download times tended to crash personal computers, even with robust institutional connections to the internet.

2. An intermediate problem with the naming of compressed folders has been corrected. Folder contents can now be extracted by basic components of common operating systems.

3. The folder names and file names now include genus names to indicate the folder and file contents. Users no longer need to browse through numerous folder contents to locate species of interest. This elegant convention obviates my suggestion that the xlxs spreadsheet could add a column that mapped specimens to the appropriate folder.

I thank the editor, Kirsten Elger, for guiding me and the authors through these successive improvements.

Most of the image files are paired. An image of the entire rock sample now includes a ruler for scale. For the close-up images of the graptolite specimens on the rock surface, the dimensions of morphologic features can readily be estimated by comparison. Earlier versions did not provide scales and this was not possible.

The Accompanying Manuscript Document
I have annotated the PDF file with edits, questions and suggestions. Many simply involve subject-verb agreement or the use of definite and indefinite articles (the, an, a). Here are some more significant issues.

1. Some species names have been emended from the original publication. Either an erroneous published identification has been corrected, or the published name has been synonymized with a previously named species, or the specimen has been re-assigned to a newer species that was established after publication of the paper to which the museum specimen is connected. The spreadsheet provides both names, emended and “tagged,” but does not distinguish between the three possible reasons to emend the name. Some of this uncertainty could be clarified if both the names included author and year. Those two terms are, after all, requirements of a valid taxon name. Readers may be unsure what the authors mean by “tagged” in this context.
2. The manuscript states that paleontologists (plural) have provided these taxonomic corrections or updates. Only one author (XM) is credited with this kind of contribution; nobody else is credited for this in the acknowledgements. It would seem that any user wishing to reference an emended name in a synonymy list would, therefore, need to credit the name to the current authors; i.e. Xu et al. 2022. No detailed justifications for the amendments are provided in the xlxs file.
3. Users of the database might wish to construct or correct range charts for the localities. They would need to know whether an emended name applies only to one specimen from that locality, to all specimens at one horizon, or to all uses of the name in the cited reference. This kind of uncertainty can arise from traditional publications too. It is difficult to update range charts, but frequently desirable.
4. Some users might not have easy access to the cited Chinese literature. Is it possible for the database to include at least the systematic taxonomy sections from those papers? I imagine there would be copyright issues.
5. The manuscript has the problem of addressing experts, amateurs and students – three stated potential users of the database with potentially very different levels of sophistication. Some parts, like the first paragraph, are not needed for experts and tend to be oversimplified for students. Other parts omit advanced considerations, like the formal treatment of synonymy and revision.

Hide

EF by Anna Mirena Feist-Polner (24 Jun 2022)

EF by Anna Mirena Feist-Polner (29 Jun 2022) Manuscript Author's tracked changes

ED: Reconsider after major revisions (22 Jul 2022) by Kirsten Elger

AR by Honghe Xu on behalf of the Authors (19 Aug 2022) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (18 Sep 2022) by Kirsten Elger

RR by Peter Sadler (12 Jan 2023)

ED: Publish subject to minor revisions (review by editor) (19 Jan 2023) by Kirsten Elger

AR by Honghe Xu on behalf of the Authors (28 Jan 2023) Author's response Author's tracked changes Manuscript

ED: Publish subject to minor revisions (review by editor) (09 Apr 2023) by Kirsten Elger

AR by Honghe Xu on behalf of the Authors (14 Apr 2023) Author's response Author's tracked changes Manuscript

ED: Publish as is (05 May 2023) by Kirsten Elger

AR by Honghe Xu on behalf of the Authors (06 May 2023)

Short summary

A multi-dimensional and integrated dataset of fossil specimens is described. The dataset potentially contributes to a range of scientific activities and provides easy access to and virtual examination of fossil specimens in a convenient and low-cost way. It will greatly benefit paleontology in research, teaching, and science communication.