the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 26 Aug 2026
PhaFoS: A Global Database of Phanerozoic Foraminiferal Size
Abstract. Body size is a key phenotypic trait of living organisms, reflecting ecological function and individual metabolism. Evolutionary changes in body size also provide information on interactions between the organisms and their changing physical and biotic environments. Despite the utilitarian value of body size for understanding evolution and past environments, we lack a standardized, occurrence-based, Phanerozoic body size database for most groups of organisms, limiting large-scale studies of size evolution and size-environment interactions. Foraminifera are an ideal group to fill this data gap, given their abundant fossil record, extensive palaeogeographic distribution, and sensitivity to environmental changes. Here, we present PhaFoS, short for Phanerozoic Foraminifera Size database. PhaFoS compiles 118,885 specimens, of which 103,121 from 963 published papers and the remaining 15,764 from our field colletions, representing 4,167 genera and 34,256 species. Each record is linked to standardized and queryable metadata fields, including test volume (log₁₀ μm³), taxonomy, stratigraphic age, and both modern and paleogeographic coordinates. This database can be used to quantify temporal and spatial patterns of foraminiferal body-size over the Phanerozoic, and to explore the body size relationships with environmental conditions and their changes. The data files described in this paper are available at https://doi.org/10.5281/zenodo.20777319 (Feng et al., 2026).
- Preprint
(1843 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 02 Nov 2026)
-
RC1: 'Comment on essd-2026-458', Rossana Martini, 01 Oct 2026
reply
-
AC1: 'Reply on RC1', Haijun Song, 02 Oct 2026
reply
We sincerely thank you for your positive feedback and valuable comments on our manuscript. We have corrected the typo in Line 66 by changing the capital “S” to a lowercase “s”.
Citation: https://doi.org/10.5194/essd-2026-458-AC1
-
AC1: 'Reply on RC1', Haijun Song, 02 Oct 2026
reply
-
RC2: 'Comment on essd-2026-458', Pedro Monarrez, 19 Oct 2026
reply
This manuscript and the foram body size database by Feng et al. are important contributions to the ongoing work on body size evolution in the fossil record and to the expansion of organismal body size data. The authors present PhaFoS, a global, specimen-based database of foram body size spanning the Phanerozoic. The database primarily consists of linear body size measurements and estimated converted volumes, each linked to taxonomic, geographic, stratigraphic, ecological, and bibliographic metadata. This database fills an important gap in fossil foram body-size data and will be a valuable resource for macroevolutionary and paleoecological research. I especially appreciate that every record can be traced to its original source, that the authors standardized taxonomic and chronostratigraphic frameworks, and that they explicitly discuss the dataset's sampling limitations, which provides room for continued growth and improvement.
Overall, this is a strong data contribution. I only have a few minor comments, most of which concern making the dataset easier to reuse. Otherwise, I believe this manuscript should be accepted once these issues have been addressed.
Minor Issues:
- Depositional environment: My biggest concern is using only two categories, shallow water and deep water, to represent depositional environments. These coarse environmental categories hide ecologically important differences in data among specific depositional environments. As someone interested in the specific depositional environments where occurrences are found, I suspect other users of this database will want to preserve these distinctions. I completely understand that environmental descriptions in the original literature vary widely in definitions and, in turn, quality, which may be why the authors chose broad categories. If this is the case, the reasoning for a binary environmental categorization should be justified in the text. Alternatively, I suggest that the authors keep the original depositional environment reported in each source and add a hierarchical standardized category with uncertainty flags, similar to what Macrostrat currently uses. This would improve the dataset's usefulness for someone interested in depositional environmental context.
- Ages and paleocoordinates: It is unclear to me how numerical geologic ages and paleocoordinates were assigned, particularly for records with broad temporal resolution. I note that a link to the paleogeographic reconstruction website (https://docs.deeptime.world) was added to the text; however, the authors need to provide more detail on how they applied this approach. I would also encourage the authors to add age and coordinate uncertainty fields so users can filter records accordingly.
- The binary SymbiontBearing YES/NO field could be problematic if “NO” sometimes means “known not to host symbionts” and sometimes means “no evidence available.” Same for ecological preference and possibly test composition.
- Metadata: The metadata would benefit from clearer vocabularies and explicit conventions for missing data. Similarly, inferred traits and volume estimates should include confidence or source fields so users can distinguish measured values from estimates.
- Heim et al. 2015; Science should be cited in the Introduction section (Line 77) where previous marine body size databases are discussed, as that is currently the largest body size database of fossil marine animals available. I see that it is cited elsewhere in the manuscript, but it belongs in that paragraph as well.
Otherwise, great and much-needed contribution!
Pedro Monarrez
Virginia TechCitation: https://doi.org/10.5194/essd-2026-458-RC2 -
AC2: 'Reply on RC2', Haijun Song, 21 Oct 2026
reply
- Response: Thank you for this thoughtful comment. We agree that the distinction between shallow-water and deep-water environments does not capture the full ecological variability among specific depositional settings. We have clarified in the revised manuscript that this broad classification was adopted to improve comparability across sources with highly heterogeneous environmental terminology and levels of detail. We have also added a statement directing users to the original sources, through the reference links provided in the database, for more detailed depositional environment descriptions where available (see Lines 272-284).
- Response: Thank you for highlighting the need for a clearer description of age assignment and palaeogeographic reconstruction.For records assigned to a standard geological stage, we now specify that numerical ages were assigned using the midpoint between the start and end ages of that stage, based on the age boundaries provide in the latest International Chronostratigraphic Chart (Cohen et al., 2025). Broader stratigraphic assignments were retained where finer resolution was not supported by sufficient evidence (see Lines 235-242); We have also clarified that palaeogeographic coordinates were reconstructed using the PALEOMAP plate reconstruction model in Time Machine. Modern latitude and longitude and the corresponding start and end ages were uploaded and mapped to the appropriate coordinate and age fields, and the resulting palaeocoordinates were incorporated into the database (see Lines 263-269).
- Response: Thank you for highlighting this important distinction. We agree that insufficient information should not be interpreted as evidence of absence. We have clarified in the revised manuscript that NO in the Symbiont Bearing field is assigned only where published information supports the absence of algal endosymbionts. Where neither the original source nor the consulted references provide sufficient information for a reliable assignment, the entry is left blank. The same convention applies to Ecological Preference and Test Composition. We now explicitly state that blank entries indicate unknown or insufficiently documented traits, rather than the absence of a trait (see Lines 293-294 and 302-306).
- Response: Thank you for this helpful suggestion. We have reviewed the metadata and added an “estimate” annotation to derived or estimated values, including reconstructed palaeogeographic coordinates, numerical ages assigned as the midpoints of the corresponding geological stages, and calculated test volumes. These annotations help users distinguish derived or estimated values from directly reported or measured values. We have also clarified the convention for missing trait data in the revised manuscript: entries are left blank where neither the original source nor the consulted references provide sufficient information for a reliable assignment. Blank entries therefore indicate unknown or insufficiently documented traits, rather than the absence of a trait(see Lines 302-306).
- Response: Thank you. We have added Heim et al. (2015) to the relevant paragraph in the Introduction(see Line 77).
-
RC3: 'Reply on AC2', Rossana Martini, 21 Oct 2026
reply
The authors have satisfactorily addressed the comments and provided useful clarifications that improve the transparency and reusability of the database. I particularly appreciate the clearer distinction between missing information and documented trait absence, as well as the identification of estimated values.
In my opinion, no further major concerns and recommend acceptance of the manuscript in its revised form.
Citation: https://doi.org/10.5194/essd-2026-458-RC3 -
AC3: 'Reply on RC3', Haijun Song, 21 Oct 2026
reply
Thank you very much for your positive comments.
Citation: https://doi.org/10.5194/essd-2026-458-AC3
-
AC3: 'Reply on RC3', Haijun Song, 21 Oct 2026
reply
Data sets
PhaFoS: A Global Database of Phanerozoic Foraminiferal Size Yan Feng and Haijun Song https://doi.org/10.5281/zenodo.20777319
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 377 | 166 | 154 | 697 | 157 | 126 |
- HTML: 377
- PDF: 166
- XML: 154
- Total: 697
- BibTeX: 157
- EndNote: 126
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This manuscript is very well structured and clearly written. The numerous figures and tables are highly appropriate and effectively support the text, greatly facilitating the understanding of the manuscript and the information presented.
In my view, the proposed workflow is particularly relevant and robust, and is appropriately based on the most widely used and established classification schemes. The six criteria selected to build the database are comprehensive and provide a solid framework for organising the data.
The taxonomic distribution within the database is presented very clearly and described in considerable detail. The accompanying figures are particularly effective, providing an excellent visual synthesis of the information discussed in the text.
The section providing instructions and recommendations for the correct use of the database is especially valuable. Importantly, it clearly outlines the limitations of the dataset while providing practical guidance on how the data should be appropriately interpreted and used. This considerably enhances the transparency and usefulness of the database for future users.
Overall, this is a carefully prepared, well-presented and valuable contribution. The manuscript provides a clear and rigorous framework, and the resulting database should represent a useful resource for the scientific community.
Given the primarily descriptive nature of this manuscript, which presents PhaFoS as a global database of Phanerozoic foraminiferal size, I have no specific scientific issues or concerns to raise. The methodology, structure and intended use of the database are clearly presented and, in my view, sufficiently well justified throughout the manuscript.
I found just one typo to correct at:
Line 66 individuals (Cope, 1896; Stanley, 1973). Furthermore, (S) size variations in individuals,