the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 23 Jun 2025
The 3D submicron-scale skeletal reconstruction of Nannoconus (Cretaceous calcareous nannofossil) – Insights on biomineralization
Abstract. Nannoconus (~5–20 μm) was a major biocarbonate producer in the Early Cretaceous seas (~150–120 Ma). The heavy calcitic skeletons (~200–1400 picogram) of this nannoplankton have contributed massive carbonate accumulations for over ~30 million years. The skeletal microstructure is characterized by an interlocking arrangement of calcitic lamellae spanned around a central canal. The biomineralization process involved in producing the sophisticated skeleton is investigated for the first time. Ptychography X-ray computed tomography (PXCT) with synchrotron radiation is applied to an isolated skeleton, to obtain a 3D set of tomographic images with ~ 40 nm spatial resolution. This 3D set was processed to virtually segment the individual calcitic lamella and reconstruct the full skeleton through constraining different lengths and angles. The lamellae are repetitively stacked in two distinct inclinations, one following the other, and producing segments combined to form the entire skeleton. Individual lamellae were calcified in a “template” of organic layer containing amino acid(s)/biomolecule(s), responsible for creating the interlocking arrangement. Our study of Nannoconus provides a simple yet potent approach to the analysis of biomineralized microstructures characterized by the repetitive arrangement of calcitic units as commonly seen in the calcareous nannoplankton.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(2699 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2699 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-1840', Anonymous Referee #1, 12 Jul 2025
-
AC1: 'Reply on RC1', Rajkumar Chowdhury, 25 Aug 2025
Dear reviewer,
We sincerely appreciate the encouraging remarks regarding this work. We also acknowledge the five points of concern raised in this study and would now like to take the opportunity to respond to these points as follows:
Comments:
- The two models in Figure 9 look qualitatively similar and it is unclear from the text why the layer model was abandoned for further discussion.
Response: The two models indeed result in similar reconstructions of Nannoconus’s skeleton, the segment model was used for the detailed discussion on the biomineralization for the following reasons:
(A) The layered structure of Nannoconus is distinctively visible in Scanning Electron Microscope (SEM) images, however, clear segment boundaries are also observed in several species. These include both one of the youngest known species, N. funiculus (reported at ~90 Ma; Lees and Bown, 2016; Fig. a, in the figure, attached as supplement), and one of the oldest, N. compressus (reported at ~140 Ma; Bralower et al., 1989; Fig. d, in the figure, attached as supplement). Moreover, in many recrystallized or overgrown specimens; where the original individual lamellae have fused into thicker, brick-like units, the segments remain, however, clearly recognizable (e.g., Fig. d, in the figure, attached as supplement). These observations indicate that the arrangement of lamellae in layers is insufficient to explain the formation of the segments in Nannoconus. In support of this interpretation, we have presented four SEM images (Figs. a-d, in the figure, attached as supplement) of four different Nannoconus species, each illustrating the persistent and clear segment boundaries across species. We will also add these images in the revised version of the manuscript. (B) As presented in the manuscript, the Genus Nannoconus belongs to the Family Nannoconaceae (Reinhardt, 1966), which is included in the Order Braarudosphaerales (Aubry, 2013; Lees and Bown, 2016). This order also includes the Family Braarudosphaeraceae, which shares a strong evolutionary link with Nannoconaceae; and therefore, with Nannoconus, as described by Lees and Bown (2016). According to Lees and Bown (2016), Braarudosphaeraceae are characterized by “five segments formed from stacks of non-imbricated laminae/elements”, whereas Nannoconaceae are defined by “numerous stacked, imbricating elements.” Here, “laminae/elements” refers to lamellae. An extant species of Braarudosphaeraceae, Braarudosphaera bigelowii, calcifies within an organic template divided into five compartments, each corresponding in shape to a segment (Hagino et al., 2016). Given the close evolutionary relationship between these two Families, it is reasonable to hypothesize that Nannoconaceae; (and therefore, Nannoconus) may have calcified via a similar process, where imbricating lamellae were stacked into segments. This inference, combined with the success of skeletal reconstruction based on the segment-model, provides strong support for a segment-based mode of calcification in Nannoconus, and justifies the continued use of this model in further analyses and discussion.
- Only one lamella is segmented in Figure 4. The authors say that the segmentation is close to the limit of the methodology. Therefore, it is important to segment several more lamellas so there is a way to assess how conserved this morphology is.
Response: The order Braarudosphaerale is defined (Aubry, 2013) as “consisting of identical, imbricated segments that are stacks of lamellae of similar shape.” As Nannoconus belongs to the same order (Aubry, 2013, 2025), it can be inferred that its lamellae are also of “similar” shape/morphology. The results of the PXCT experiment of N. globulus, shows that the lamellae in both base and apex have the same morphology. Based on these arguments, we concluded that the skeleton of N. globulus is composed of morphologically similar lamellae.
- When constructing the full shell model in silico, is there a condition that each voxel is hosting only a single lamella? In other words, is physical overlap of two lamella in the same volume avoided?
Response: The physical overlap between two consecutive lamellae is effectively close to zero. After several trials the values of different angles and lengths are taken such that the lamellae merely touch each other without any physical overlap. Therefore, a single voxel contains only one lamella.
- The geometrical descriptions are fundamental for the study and the authors try their best to explain and define all aspects, nevertheless, the terminology is difficult to follow. If the authors could improve the visualization of the angles in Figures 5 and 3 it can make this aspect clearer.
Response: We understand the inherent difficulty for the comprehension of all the geometric parameters specifically in the 3D skeleton. We have taken note of this point and will try to improve the 3D visualization of the angles. Specifically, we plan to provide additional images of the skeleton in oblique and cross-sectional views where both the angles (i.e., inclination and tilt) will be simultaneously presented for better clarity.
- It is very difficult to follow the discussion of B. bigelowii on page 22 and to understand which similarities the authors propose. A visualization of this structure can help.
Response: In page 22, a hypothesis on the biomineralization of Nannoconus’s skeleton as a combination of segments, templated by an organic substrate is proposed in comparison to the biomineralization of B. bigelowii as proposed by Hagino et al., (2016). We will add an image of B. bigelowii (modifying from Hagino et al., 2016), in the appendices, with the segments and the organic substrate well-demarcated, for the clarification of its structural similarities with the skeleton of Nannoconus.
Rajkumar Chowdhury,
On behalf of all the co-authors.
Reference cited:
Aubry, M.P.: Cenozoic Coccolithophores: Braarudosphaerales, Atlas of Micropaleontology series Micropaleontology Press, 540 New York (336 pp), 2013.
Aubry, M.P.: Biomineralization in the Calcareous Nannoplankton Phenotypic Expressions Across Life Cycles, Geometric Control on Diversification, and Origin, Minerals, 15, 322, 2025.
Bralower, T. J., Monechi, S., and Thierstein, H. R.: Calcareous nannofossil zonation of the Jurassic-Cretaceous boundary interval and correlation with the geomagnetic polarity timescale, Marine Micropaleontology, 14, 153–235, 1989.
Hagino, K., Tomioka, N., Young, J. R., Takano, Y., Onuma, R., and Horiguchi, T.: Extracellular calcification of Braarudosphaera bigelowii deduced from electron microscopic observations of cell surface structure and elemental composition of pentaliths, Marine Micropaleontology, 125, 85–94, https://doi.org/10.1016/j.marmicro.2016.04.002, 2016.
Lees, J. A. and Bown, P. R.: New and intriguing calcareous nannofossils from the Turonian (Upper Cretaceous) of Tanzania, Journal of Nannoplankton Research, 36, 83–95, 2016.
Reinhardt, P.: Zur Taxionomie und Biostratigraphie des fossilen Nannoplanktons aus dem Malm, der Kreide und dem Alttertiär Mitteleuropas: mit 1 Tabelle, PhD Thesis, Dt. Verlag für Grundstoffindustrie, 1966.
-
AC1: 'Reply on RC1', Rajkumar Chowdhury, 25 Aug 2025
-
RC2: 'Comment on egusphere-2025-1840', Jeremy Young & Angela Fraguas (co-review team), 16 Jul 2025
Review of Chowdury et al. “The 3D submicron-scale skeletal reconstruction of Nannoconus
(Cretaceous calcareous nannofossil) - Insights on biomineralization”
This paper is based on a ground-breaking investigation of Nannoconus using Ptychographic X-ray computed tomography (PXCT). This is a synchrotron-based approach which has allowed reconstruction of the complex structure of these calcareous nannofossils at sub-micron level. This is the first application of a 3D imaging technique of this type to investigate the shape, size and disposition of the individual elements in addition to the gross morphology of the object. The technique overcomes some of the limits of previous SEM and LM based methods. Nannoconus is a good choice of subject since their skeletal elements, nannoconids, were important rock-formers in the Early Cretaceous and since their evolutionary relationship to coccolithophores and other nannofossils has been uncertain.
Nannoconus is formed of numerous wedge-shaped laminae arranged in spiral layers showing clockwise imbrication in side view – i.e. left-handed helices.
van Niel (1992, 1994) observed that alternate layers of elements were either arranged sub-parallel to the spiral surfaces with only slight overlap (type A plates), or were significantly inclined relative to this surface and so overlapped giving an imbricate structure to those layers (type B plates). These alternate layers spiral around the nannolith, but it has not been established how many helical layers – i.e whether there is one set of A-B layer pairs, two, or several. There is also a suggestion (van Niel 1992, Young et al. 1997) that the two cycles have different crystallographic orientations and that this causes the dark spiral lines clearly visible in polarising light microscopy. This suggestion is unproven with the alternative being that the entire nannolith is formed of elements with sub-tangential orientation and that the dark lines are some type of interference effect.
Aubry (2013, 2025) argued that plates in successive players overlaid each other regularly and so nannoconids could be considered as being formed of segments - as also hinted at by Bronnimann (1955). This segment model is significant since it suggests closer affinity to Braarudosphaeraceae which are formed of 5 clearly separated segments with a laminar sub-structure. This is an intriguing suggestion, but it has not been unambiguously demonstrated. Studies of Late Jurassic nannofossils (Bergen et al. 2014, Varol & Bowman 2019) have revealed possible common ancestors for Nannoconus and the Braarudosphaeraceae strengthening the case for the segment model.
Given this, PXCT investigation of Nannoconus seems extremely promising as a way of resolving uncertainties of the structure, as well as paving the way for applying the technique to other nannofossils. However, although it is stated that the technique has been successfully applied to 5 hand-picked, well-preserved, specimens, only very limited results are presented here. The process of segmenting out a lamella from the PXCT data is shown and I would have expected this approach to be applied to the entire nannolith, as is routinely done with CT reconstructions of larger fossils. This should have allowed direct testing of the segment model, determination of the number of helical spirals, etc.
However, no such reconstruction is provided, instead the bulk of the results section is concerned with mathematical modelling of possible Nannoconus structures based on the segmentation of a single lamella from one PXCT image stack. This modelling apparently showed that Nannoconus-like structures can be created using the segment model. The modelling approach is however, as noted by other reviewers, hard to follow. In part this may be a result of inherent complexity and of mperfect presentation, but some aspects give cause for concern. Notably (1) The layer and segment models appear to be treated as mutually exclusive alternatives which is clearly illogical since both models can be valid descriptions of the same structure. By analogy, a stretcher bond brick wall can be considered as being formed of layers of bricks, but can equally validly be described as formed of columns of bricks with alternating offsets between successive bricks, or as formed of bricks arranged in sloping rows. These three ways of describing the arrangement of the bricks are each valid and are not mutually exclusive. In the same way the layer and segment descriptions of Nannoconus structure can both be correct and are not mutually exclusive. The layer structure is readily observable in electron micrographs and is clearly correct. The segment model is less obvious, but the data presented here provides evidence in favour of it. (2) In many places in section 4.3 layers/segments are referred to – e.g. “the total number (N) of layers/segments in the whole skeleton of N. globulus is calculated as 12.” This seems to indicate some confusion, layers and segments are different subdivisions of the Nannoconus structure so a layer/segment has no obvious meaning, and there is no obvious reason why the number of layers should be the same as the number of segments. (3) Twin lamellae – in section 4.1 it is stated that “Lamella-A + lamella-B formed twin lamellae.” It is not clear what this statement means, in particular is it implied that the A and B lamellae are crystallographic twins?
Conclusion: Although inspired by the PXCT study, the primary focus of this paper is mathematical modelling of possible Nannoconus structures. I recommend that the presentation of this modelling is carefully revised and the conclusions that can be drawn from it clarified. I would also welcome a detailed presentation of the results of PXCT study, either here or in a separate paper.
References cited
Aubry, M. -P. (2013). Cenozoic Coccolithophores: Braarudosphaerales. Micropaleontology Press, American Museum of Natural History, New York. 1-336.
Aubry, M. -P. (2025). Biomineralization in the calcareous nannoplankton phenotypic expressions across life cycles, geometric control on diversification, and origin. Minerals. 15: 1-49.
Bergen, J. A., Boesiger, T. M. & Pospichal, J. J. (2014). Low-latitude Oxfordian to Early Berriasian nannofossil biostratigraphy and its application to the subsurface of Eastern Texas. In, Hammes, U. & Gale, J. (eds) Geology of the Haynesville Gas Shale in East Texas and West Louisiana, U.S.A. American Association of Petroleum Geologists, Memoirs . 105: 69-102.
Brönnimann, P. (1955). Microfossils incertae sedis from the Upper Jurassic and Lower Cretaceous of Cuba. Micropaleontology. 1(1): 28-51
van Niel, B. E. (1992). New observations on the morphology of Nannoconus. In, Hamrsmid, B. & Young, J. R. (eds) Nannoplankton Research, Proceedings of the 4th INA Conference, Prague 1991, vol II. Knihovnicka ZPN . 14a: 73-85.
van Niel, B. E. (1994a). A review of the terminology used to describe the genus Nannoconus (calcareous nannofossil, incertae sedis). Cahiers de Micropaléontologie. 9: 27-47
Varol, O. & Bowman, A. R. (2019). Taxonomic revision of selected Late Jurassic (Tithonian) calcareous nannofossils and the application of mobile mounting. Neues Jahrbuch für Geologie und Paläontologie, Abhandlungen. 291: 65-87.
Young, J. R., et al. (1997). Guidelines for coccolith and calcareous nannofossil terminology. Palaeontology. 40: 875-912.
Citation: https://doi.org/10.5194/egusphere-2025-1840-RC2 -
AC2: 'Reply on RC2', Rajkumar Chowdhury, 25 Aug 2025
Dear reviewers,
We greatly value the comments and concerns regarding our investigation of Ptychographic X-ray computed tomography (PXCT) as applied to Nannoconus, a calcareous nannofossil, that was the main bicarbonate producer of Early Cretaceous oceans. One of the goals of this paper is to provide insights into the biomineralization process of the Nannoconus’s microskeleton, previously unknown, through a definitive understanding of its 3D skeletal microstructure.
In this regard, we have successfully applied the PXCT technique on five hand-picked specimens of Nannoconus. Among them, Nannoconus globulus was chosen for detailed explanation and discussion, because of its well-defined globular morphology and clearly distinguishable lamellar inclinations. Although the modelling approach is demonstrated using N. globulus, it is adaptable, by modifying the radius, to reproduce different morphologies belonging to various species of Nannoconus, as illustrated in Figure 11 in the manuscript. Computed tomographic reconstruction (CT) which is often applied to larger nannofossils, generally combines with image segmentation (Segmentation refers to the process of extracting features from tomographic image stack and should not be confused with the term “segment”, used to describe the 3D microstructure of Nannoconus). Common segmentation methods include watershed segmentation or U-Net based segmentation (Reznikov et al., 2020). These segmentation methods could not be applied to Nannoconus, because of its small size limiting to the resolution achieved in our experiment. As a result, we have manually segmented a “lamella”, the basic structural unit and used it for reconstructing the full skeleton. However, this manual segmentation approach is not applicable to the entire Nannoconus because: (A) It is extremely time consuming to manually segment numerous lamellae of Nannoconus, from ~300 tomographic image slices. (B) The lamellae frequently overlap with each other, due to post-depositional overgrowth (a process that dissolves and reprecipitates the calcite), making them difficult to distinguish for segmentation. Hence, we have segmented a distinctly identifiable lamella and utilized it as a unit to reconstruct the layers, alternatively the segments and finally the entire skeleton. The number of layers and segments, along with various angle and length parameters, were optimized through several trials to ensure that the reconstructed skeleton closely resembles the actual specimen. This unit-based modelling strategy, combined with optimization of various parameters of key angles (i.e., inclination and tilt) and lengths (i.e., radius), thus offers a practical framework for physically reconstructing and studying other micrometric nannofossil skeletons.
We would now like to take the opportunity to respond to the three points of concern as noted.
(1) In this study, the 3D microstructural arrangement of the Nannoconus’s skeleton is described by two different ways notably, the layer model (following van Niel, 1993) and segment model (following Aubry, 2013, 2025) and thus both of them were indeed taken as valid descriptions of the skeleton for the reconstruction. While the two models yield comparable reconstructions of the Nannoconus skeleton, the segment model however, was chosen to further hypothesize insights into skeletal biomineralization, aligning with the goal of this paper, for the following reasons:
(A) Even though the layered structure of Nannoconus is clearly visible in Scanning Electron Microscopic (SEM) images, segment boundaries are also distinctly observed in several species, including in one of the youngest species (N. funiculus, reported ~90 Ma, Lees and Bown, 2016, Fig. a, in the figure, attached as supplement) and one of the oldest species (N. compressus, reported ~140 Ma, Bralower et al., 1989; Fig. d, in the figure, attached as supplement). In addition, in many recrystallized/overgrown Nannoconus in which the individual lamellae are no longer visible (they have fused to form sorts of bricks; i.e., became thicker) the organization in segments remains clear (Fig. d, in the figure, attached as supplement). Therefore, a sole arrangement of the lamellae in layers to form the entire skeleton is not sufficient to explain the individualization of segments. To support this observation, we have presented four Scanning Electron Microscopic images (Figs. a-d, in the figure, attached as supplement) of these species that highlight the segment boundaries and will also add them in the revised version of the paper. (B) As noted in this paper, the Genus Nannoconus belongs to the Family Nannoconaceae (Reinhardt, 1966), included in the Order Braarudosphaerales (Aubry, 2013; Lees and Bown, 2016). This order also includes the family Braarudosphaeraceae, which has a strong evolutionary link to Nannoconaceae; and thus, to Nannoconus; as described by Lees and Bown (2016). According to Lees and Bown (2016); Braarudosphaeraceae is composed of “five segments formed from stacks of non-imbricated laminae/elements” and Nannoconaceae is composed of “numerous stacked, imbricating elements”. Laminae/elements refer to lamellae. A living species Braarudosphaera bigelowii, a member of the Family Braarudosphaeraceae, calcifies in an organic substrate, divided into five parts, with each of the parts mimicking the shape of a segment (Hagino et al., 2016). Considering the strong evolutionary link between the two Families, it is reasonable to infer that the Nannoconaceae (and therefore Nannoconus), may have calcified in a similar process of stacking the imbricating lamellae into segments. Therefore, combined with the successful skeletal reconstruction from the segment model, an imperative segment-based calcification of Nannoconus, strengthens the obvious choice of the segment model for further discussion. (C) It is also worth noting that the boundaries of segments may become less distinct at higher angles of rotation (θ) of segments (defined in subsection 4.3.2), as explained in Figure 10 of the manuscript. This effect could potentially obscure the visibility of segment boundaries in electron microscopy making them hard to recognize in several species.
(2) In subsections 2.2.1 and 2.2.2, we have calculated the total number of layers and segments from the SEM image of N. globulus. Both the number of layers and the number of segments were determined to be 12. Afterwards, this value is retained for the skeletal reconstruction presented in section 4.3, depending on the chosen model (i.e., the layer model or the segment model).
(3) The term “twin lamellae” is used here to describe the repeating pair of lamellae (i.e., lamella-A and lamella-B) that together construct the full skeleton of Nannoconus. This term does not refer to crystallographic twinning, but rather to the microstructural arrangement and alternating inclinations of the lamellae. We acknowledge the ambiguity in this terminology and will provide a clearer explanation in the revised version.
In conclusion, we acknowledge that PXCT alone is not sufficient to definitively distinguish between the two models. However, it indeed supports the successful 3D reconstruction of the Nannoconus skeletal structure. When combined with existing 2D SEM observation, as well as biomineralization hypotheses, the evidence supports the interpretation of the Nannoconus's skeleton as being composed of “segments”, in the same way as Braarudosphaera is and other Mesozoic nannofossils (Aubry 2025 and forthcoming). We will clarify this point in the revised manuscript, along with an improved presentation of the geometric parameters and clear description of the ambiguous terms.
Rajkumar Chowdhury,
On behalf of all the co-authors,
Reference cited:
Aubry, M.P.: Cenozoic Coccolithophores: Braarudosphaerales, Atlas of Micropaleontology series Micropaleontology Press, 540 New York (336 pp), 2013.
Aubry, M.P.: Biomineralization in the Calcareous Nannoplankton Phenotypic Expressions Across Life Cycles, Geometric Control on Diversification, and Origin, Minerals, 15, 322, 2025.
Aubry, M.P.: Micalithophores: Braarudosphaerales, SEPM Concepts in Sedimentology and Paleontology, CSP XX, forthcoming, 2026.
Bralower, T. J., Monechi, S., and Thierstein, H. R.: Calcareous nannofossil zonation of the Jurassic-Cretaceous boundary interval and correlation with the geomagnetic polarity timescale, Marine Micropaleontology, 14, 153–235, 1989.
Hagino, K., Tomioka, N., Young, J. R., Takano, Y., Onuma, R., and Horiguchi, T.: Extracellular calcification of Braarudosphaera bigelowii deduced from electron microscopic observations of cell surface structure and elemental composition of pentaliths, Marine Micropaleontology, 125, 85–94, https://doi.org/10.1016/j.marmicro.2016.04.002, 2016.
Lees, J. A. and Bown, P. R.: New and intriguing calcareous nannofossils from the Turonian (Upper Cretaceous) of Tanzania, Journal of Nannoplankton Research, 36, 83–95, 2016.
Reinhardt, P.: Zur Taxionomie und Biostratigraphie des fossilen Nannoplanktons aus dem Malm, der Kreide und dem Alttertiär Mitteleuropas: mit 1 Tabelle, PhD Thesis, Dt. Verlag für Grundstoffindustrie, 1966.
Reznikov, N., Buss, D. J., Provencher, B., McKee, M. D., and Piché, N.: Deep learning for 3D imaging and image analysis in biomineralization research, Journal of Structural Biology, 212, 107598, 2020.
Van Niel, B. E.: Early CretaceousNannoconus’(calcareous nannofossil, incertae sedis) in NW Europe, University of London, University College London (United Kingdom), 1993.
-
AC2: 'Reply on RC2', Rajkumar Chowdhury, 25 Aug 2025
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-1840', Anonymous Referee #1, 12 Jul 2025
The study “The 3D submicron-scale skeletal reconstruction of Nannoconus (Cretaceous calcareous nannofossil) - Insights on biomineralization” by Chowdhury et al. reports on the 3D imaging of Nannoconus shell, its segmentation into building units, and their computational reconstruction into a model shell. The work is very interesting as it gives the first 3D view into the abundant Nannoconus morphology observed in sediments. This brings an important addition to other structures of nanoplankton (most studied are probably the coccolithophorids). I very much appreciate the effort to use the segment morphology of the basic unit and to re-build from it in silico the full shell. The limitations of the study emerge from the fact that there is no living reference for the shells, and several assumptions are being made along the way. Nevertheless, this is a beautiful study that deserves publication after the following comments are addressed.
Comments:
- The two models in Figure 9 look qualitatively similar and it is unclear from the text why the layer model was abandoned for further discussion.
- Only one lamella is segmented in Figure 4. The authors say that the segmentation is close to the limit of the methodology. Therefore, it is important to segment several more lamellas so there is a way to assess how conserved is this morphology.
- When constructing the full shell model in silico, is there a condition that each voxel is hosting only a single lamella? In other words, is physical overlap of two lamella in the same volume avoided?
- The geometrical descriptions are fundamental for the study and the authors try their best to explain and define all aspects, nevertheless, the terminology is difficult to follow. If the authors could improve the visualization of the angles in Figures 5 and 3 it can make this aspect clearer.
- It is very difficult to follow the discussion of B. bigelowii on page 22 and to understand which similarities the authors propose. A visualization of this structure can help.
Citation: https://doi.org/10.5194/egusphere-2025-1840-RC1 -
AC1: 'Reply on RC1', Rajkumar Chowdhury, 25 Aug 2025
Dear reviewer,
We sincerely appreciate the encouraging remarks regarding this work. We also acknowledge the five points of concern raised in this study and would now like to take the opportunity to respond to these points as follows:
Comments:
- The two models in Figure 9 look qualitatively similar and it is unclear from the text why the layer model was abandoned for further discussion.
Response: The two models indeed result in similar reconstructions of Nannoconus’s skeleton, the segment model was used for the detailed discussion on the biomineralization for the following reasons:
(A) The layered structure of Nannoconus is distinctively visible in Scanning Electron Microscope (SEM) images, however, clear segment boundaries are also observed in several species. These include both one of the youngest known species, N. funiculus (reported at ~90 Ma; Lees and Bown, 2016; Fig. a, in the figure, attached as supplement), and one of the oldest, N. compressus (reported at ~140 Ma; Bralower et al., 1989; Fig. d, in the figure, attached as supplement). Moreover, in many recrystallized or overgrown specimens; where the original individual lamellae have fused into thicker, brick-like units, the segments remain, however, clearly recognizable (e.g., Fig. d, in the figure, attached as supplement). These observations indicate that the arrangement of lamellae in layers is insufficient to explain the formation of the segments in Nannoconus. In support of this interpretation, we have presented four SEM images (Figs. a-d, in the figure, attached as supplement) of four different Nannoconus species, each illustrating the persistent and clear segment boundaries across species. We will also add these images in the revised version of the manuscript. (B) As presented in the manuscript, the Genus Nannoconus belongs to the Family Nannoconaceae (Reinhardt, 1966), which is included in the Order Braarudosphaerales (Aubry, 2013; Lees and Bown, 2016). This order also includes the Family Braarudosphaeraceae, which shares a strong evolutionary link with Nannoconaceae; and therefore, with Nannoconus, as described by Lees and Bown (2016). According to Lees and Bown (2016), Braarudosphaeraceae are characterized by “five segments formed from stacks of non-imbricated laminae/elements”, whereas Nannoconaceae are defined by “numerous stacked, imbricating elements.” Here, “laminae/elements” refers to lamellae. An extant species of Braarudosphaeraceae, Braarudosphaera bigelowii, calcifies within an organic template divided into five compartments, each corresponding in shape to a segment (Hagino et al., 2016). Given the close evolutionary relationship between these two Families, it is reasonable to hypothesize that Nannoconaceae; (and therefore, Nannoconus) may have calcified via a similar process, where imbricating lamellae were stacked into segments. This inference, combined with the success of skeletal reconstruction based on the segment-model, provides strong support for a segment-based mode of calcification in Nannoconus, and justifies the continued use of this model in further analyses and discussion.
- Only one lamella is segmented in Figure 4. The authors say that the segmentation is close to the limit of the methodology. Therefore, it is important to segment several more lamellas so there is a way to assess how conserved this morphology is.
Response: The order Braarudosphaerale is defined (Aubry, 2013) as “consisting of identical, imbricated segments that are stacks of lamellae of similar shape.” As Nannoconus belongs to the same order (Aubry, 2013, 2025), it can be inferred that its lamellae are also of “similar” shape/morphology. The results of the PXCT experiment of N. globulus, shows that the lamellae in both base and apex have the same morphology. Based on these arguments, we concluded that the skeleton of N. globulus is composed of morphologically similar lamellae.
- When constructing the full shell model in silico, is there a condition that each voxel is hosting only a single lamella? In other words, is physical overlap of two lamella in the same volume avoided?
Response: The physical overlap between two consecutive lamellae is effectively close to zero. After several trials the values of different angles and lengths are taken such that the lamellae merely touch each other without any physical overlap. Therefore, a single voxel contains only one lamella.
- The geometrical descriptions are fundamental for the study and the authors try their best to explain and define all aspects, nevertheless, the terminology is difficult to follow. If the authors could improve the visualization of the angles in Figures 5 and 3 it can make this aspect clearer.
Response: We understand the inherent difficulty for the comprehension of all the geometric parameters specifically in the 3D skeleton. We have taken note of this point and will try to improve the 3D visualization of the angles. Specifically, we plan to provide additional images of the skeleton in oblique and cross-sectional views where both the angles (i.e., inclination and tilt) will be simultaneously presented for better clarity.
- It is very difficult to follow the discussion of B. bigelowii on page 22 and to understand which similarities the authors propose. A visualization of this structure can help.
Response: In page 22, a hypothesis on the biomineralization of Nannoconus’s skeleton as a combination of segments, templated by an organic substrate is proposed in comparison to the biomineralization of B. bigelowii as proposed by Hagino et al., (2016). We will add an image of B. bigelowii (modifying from Hagino et al., 2016), in the appendices, with the segments and the organic substrate well-demarcated, for the clarification of its structural similarities with the skeleton of Nannoconus.
Rajkumar Chowdhury,
On behalf of all the co-authors.
Reference cited:
Aubry, M.P.: Cenozoic Coccolithophores: Braarudosphaerales, Atlas of Micropaleontology series Micropaleontology Press, 540 New York (336 pp), 2013.
Aubry, M.P.: Biomineralization in the Calcareous Nannoplankton Phenotypic Expressions Across Life Cycles, Geometric Control on Diversification, and Origin, Minerals, 15, 322, 2025.
Bralower, T. J., Monechi, S., and Thierstein, H. R.: Calcareous nannofossil zonation of the Jurassic-Cretaceous boundary interval and correlation with the geomagnetic polarity timescale, Marine Micropaleontology, 14, 153–235, 1989.
Hagino, K., Tomioka, N., Young, J. R., Takano, Y., Onuma, R., and Horiguchi, T.: Extracellular calcification of Braarudosphaera bigelowii deduced from electron microscopic observations of cell surface structure and elemental composition of pentaliths, Marine Micropaleontology, 125, 85–94, https://doi.org/10.1016/j.marmicro.2016.04.002, 2016.
Lees, J. A. and Bown, P. R.: New and intriguing calcareous nannofossils from the Turonian (Upper Cretaceous) of Tanzania, Journal of Nannoplankton Research, 36, 83–95, 2016.
Reinhardt, P.: Zur Taxionomie und Biostratigraphie des fossilen Nannoplanktons aus dem Malm, der Kreide und dem Alttertiär Mitteleuropas: mit 1 Tabelle, PhD Thesis, Dt. Verlag für Grundstoffindustrie, 1966.
-
RC2: 'Comment on egusphere-2025-1840', Jeremy Young & Angela Fraguas (co-review team), 16 Jul 2025
Review of Chowdury et al. “The 3D submicron-scale skeletal reconstruction of Nannoconus
(Cretaceous calcareous nannofossil) - Insights on biomineralization”
This paper is based on a ground-breaking investigation of Nannoconus using Ptychographic X-ray computed tomography (PXCT). This is a synchrotron-based approach which has allowed reconstruction of the complex structure of these calcareous nannofossils at sub-micron level. This is the first application of a 3D imaging technique of this type to investigate the shape, size and disposition of the individual elements in addition to the gross morphology of the object. The technique overcomes some of the limits of previous SEM and LM based methods. Nannoconus is a good choice of subject since their skeletal elements, nannoconids, were important rock-formers in the Early Cretaceous and since their evolutionary relationship to coccolithophores and other nannofossils has been uncertain.
Nannoconus is formed of numerous wedge-shaped laminae arranged in spiral layers showing clockwise imbrication in side view – i.e. left-handed helices.
van Niel (1992, 1994) observed that alternate layers of elements were either arranged sub-parallel to the spiral surfaces with only slight overlap (type A plates), or were significantly inclined relative to this surface and so overlapped giving an imbricate structure to those layers (type B plates). These alternate layers spiral around the nannolith, but it has not been established how many helical layers – i.e whether there is one set of A-B layer pairs, two, or several. There is also a suggestion (van Niel 1992, Young et al. 1997) that the two cycles have different crystallographic orientations and that this causes the dark spiral lines clearly visible in polarising light microscopy. This suggestion is unproven with the alternative being that the entire nannolith is formed of elements with sub-tangential orientation and that the dark lines are some type of interference effect.
Aubry (2013, 2025) argued that plates in successive players overlaid each other regularly and so nannoconids could be considered as being formed of segments - as also hinted at by Bronnimann (1955). This segment model is significant since it suggests closer affinity to Braarudosphaeraceae which are formed of 5 clearly separated segments with a laminar sub-structure. This is an intriguing suggestion, but it has not been unambiguously demonstrated. Studies of Late Jurassic nannofossils (Bergen et al. 2014, Varol & Bowman 2019) have revealed possible common ancestors for Nannoconus and the Braarudosphaeraceae strengthening the case for the segment model.
Given this, PXCT investigation of Nannoconus seems extremely promising as a way of resolving uncertainties of the structure, as well as paving the way for applying the technique to other nannofossils. However, although it is stated that the technique has been successfully applied to 5 hand-picked, well-preserved, specimens, only very limited results are presented here. The process of segmenting out a lamella from the PXCT data is shown and I would have expected this approach to be applied to the entire nannolith, as is routinely done with CT reconstructions of larger fossils. This should have allowed direct testing of the segment model, determination of the number of helical spirals, etc.
However, no such reconstruction is provided, instead the bulk of the results section is concerned with mathematical modelling of possible Nannoconus structures based on the segmentation of a single lamella from one PXCT image stack. This modelling apparently showed that Nannoconus-like structures can be created using the segment model. The modelling approach is however, as noted by other reviewers, hard to follow. In part this may be a result of inherent complexity and of mperfect presentation, but some aspects give cause for concern. Notably (1) The layer and segment models appear to be treated as mutually exclusive alternatives which is clearly illogical since both models can be valid descriptions of the same structure. By analogy, a stretcher bond brick wall can be considered as being formed of layers of bricks, but can equally validly be described as formed of columns of bricks with alternating offsets between successive bricks, or as formed of bricks arranged in sloping rows. These three ways of describing the arrangement of the bricks are each valid and are not mutually exclusive. In the same way the layer and segment descriptions of Nannoconus structure can both be correct and are not mutually exclusive. The layer structure is readily observable in electron micrographs and is clearly correct. The segment model is less obvious, but the data presented here provides evidence in favour of it. (2) In many places in section 4.3 layers/segments are referred to – e.g. “the total number (N) of layers/segments in the whole skeleton of N. globulus is calculated as 12.” This seems to indicate some confusion, layers and segments are different subdivisions of the Nannoconus structure so a layer/segment has no obvious meaning, and there is no obvious reason why the number of layers should be the same as the number of segments. (3) Twin lamellae – in section 4.1 it is stated that “Lamella-A + lamella-B formed twin lamellae.” It is not clear what this statement means, in particular is it implied that the A and B lamellae are crystallographic twins?
Conclusion: Although inspired by the PXCT study, the primary focus of this paper is mathematical modelling of possible Nannoconus structures. I recommend that the presentation of this modelling is carefully revised and the conclusions that can be drawn from it clarified. I would also welcome a detailed presentation of the results of PXCT study, either here or in a separate paper.
References cited
Aubry, M. -P. (2013). Cenozoic Coccolithophores: Braarudosphaerales. Micropaleontology Press, American Museum of Natural History, New York. 1-336.
Aubry, M. -P. (2025). Biomineralization in the calcareous nannoplankton phenotypic expressions across life cycles, geometric control on diversification, and origin. Minerals. 15: 1-49.
Bergen, J. A., Boesiger, T. M. & Pospichal, J. J. (2014). Low-latitude Oxfordian to Early Berriasian nannofossil biostratigraphy and its application to the subsurface of Eastern Texas. In, Hammes, U. & Gale, J. (eds) Geology of the Haynesville Gas Shale in East Texas and West Louisiana, U.S.A. American Association of Petroleum Geologists, Memoirs . 105: 69-102.
Brönnimann, P. (1955). Microfossils incertae sedis from the Upper Jurassic and Lower Cretaceous of Cuba. Micropaleontology. 1(1): 28-51
van Niel, B. E. (1992). New observations on the morphology of Nannoconus. In, Hamrsmid, B. & Young, J. R. (eds) Nannoplankton Research, Proceedings of the 4th INA Conference, Prague 1991, vol II. Knihovnicka ZPN . 14a: 73-85.
van Niel, B. E. (1994a). A review of the terminology used to describe the genus Nannoconus (calcareous nannofossil, incertae sedis). Cahiers de Micropaléontologie. 9: 27-47
Varol, O. & Bowman, A. R. (2019). Taxonomic revision of selected Late Jurassic (Tithonian) calcareous nannofossils and the application of mobile mounting. Neues Jahrbuch für Geologie und Paläontologie, Abhandlungen. 291: 65-87.
Young, J. R., et al. (1997). Guidelines for coccolith and calcareous nannofossil terminology. Palaeontology. 40: 875-912.
Citation: https://doi.org/10.5194/egusphere-2025-1840-RC2 -
AC2: 'Reply on RC2', Rajkumar Chowdhury, 25 Aug 2025
Dear reviewers,
We greatly value the comments and concerns regarding our investigation of Ptychographic X-ray computed tomography (PXCT) as applied to Nannoconus, a calcareous nannofossil, that was the main bicarbonate producer of Early Cretaceous oceans. One of the goals of this paper is to provide insights into the biomineralization process of the Nannoconus’s microskeleton, previously unknown, through a definitive understanding of its 3D skeletal microstructure.
In this regard, we have successfully applied the PXCT technique on five hand-picked specimens of Nannoconus. Among them, Nannoconus globulus was chosen for detailed explanation and discussion, because of its well-defined globular morphology and clearly distinguishable lamellar inclinations. Although the modelling approach is demonstrated using N. globulus, it is adaptable, by modifying the radius, to reproduce different morphologies belonging to various species of Nannoconus, as illustrated in Figure 11 in the manuscript. Computed tomographic reconstruction (CT) which is often applied to larger nannofossils, generally combines with image segmentation (Segmentation refers to the process of extracting features from tomographic image stack and should not be confused with the term “segment”, used to describe the 3D microstructure of Nannoconus). Common segmentation methods include watershed segmentation or U-Net based segmentation (Reznikov et al., 2020). These segmentation methods could not be applied to Nannoconus, because of its small size limiting to the resolution achieved in our experiment. As a result, we have manually segmented a “lamella”, the basic structural unit and used it for reconstructing the full skeleton. However, this manual segmentation approach is not applicable to the entire Nannoconus because: (A) It is extremely time consuming to manually segment numerous lamellae of Nannoconus, from ~300 tomographic image slices. (B) The lamellae frequently overlap with each other, due to post-depositional overgrowth (a process that dissolves and reprecipitates the calcite), making them difficult to distinguish for segmentation. Hence, we have segmented a distinctly identifiable lamella and utilized it as a unit to reconstruct the layers, alternatively the segments and finally the entire skeleton. The number of layers and segments, along with various angle and length parameters, were optimized through several trials to ensure that the reconstructed skeleton closely resembles the actual specimen. This unit-based modelling strategy, combined with optimization of various parameters of key angles (i.e., inclination and tilt) and lengths (i.e., radius), thus offers a practical framework for physically reconstructing and studying other micrometric nannofossil skeletons.
We would now like to take the opportunity to respond to the three points of concern as noted.
(1) In this study, the 3D microstructural arrangement of the Nannoconus’s skeleton is described by two different ways notably, the layer model (following van Niel, 1993) and segment model (following Aubry, 2013, 2025) and thus both of them were indeed taken as valid descriptions of the skeleton for the reconstruction. While the two models yield comparable reconstructions of the Nannoconus skeleton, the segment model however, was chosen to further hypothesize insights into skeletal biomineralization, aligning with the goal of this paper, for the following reasons:
(A) Even though the layered structure of Nannoconus is clearly visible in Scanning Electron Microscopic (SEM) images, segment boundaries are also distinctly observed in several species, including in one of the youngest species (N. funiculus, reported ~90 Ma, Lees and Bown, 2016, Fig. a, in the figure, attached as supplement) and one of the oldest species (N. compressus, reported ~140 Ma, Bralower et al., 1989; Fig. d, in the figure, attached as supplement). In addition, in many recrystallized/overgrown Nannoconus in which the individual lamellae are no longer visible (they have fused to form sorts of bricks; i.e., became thicker) the organization in segments remains clear (Fig. d, in the figure, attached as supplement). Therefore, a sole arrangement of the lamellae in layers to form the entire skeleton is not sufficient to explain the individualization of segments. To support this observation, we have presented four Scanning Electron Microscopic images (Figs. a-d, in the figure, attached as supplement) of these species that highlight the segment boundaries and will also add them in the revised version of the paper. (B) As noted in this paper, the Genus Nannoconus belongs to the Family Nannoconaceae (Reinhardt, 1966), included in the Order Braarudosphaerales (Aubry, 2013; Lees and Bown, 2016). This order also includes the family Braarudosphaeraceae, which has a strong evolutionary link to Nannoconaceae; and thus, to Nannoconus; as described by Lees and Bown (2016). According to Lees and Bown (2016); Braarudosphaeraceae is composed of “five segments formed from stacks of non-imbricated laminae/elements” and Nannoconaceae is composed of “numerous stacked, imbricating elements”. Laminae/elements refer to lamellae. A living species Braarudosphaera bigelowii, a member of the Family Braarudosphaeraceae, calcifies in an organic substrate, divided into five parts, with each of the parts mimicking the shape of a segment (Hagino et al., 2016). Considering the strong evolutionary link between the two Families, it is reasonable to infer that the Nannoconaceae (and therefore Nannoconus), may have calcified in a similar process of stacking the imbricating lamellae into segments. Therefore, combined with the successful skeletal reconstruction from the segment model, an imperative segment-based calcification of Nannoconus, strengthens the obvious choice of the segment model for further discussion. (C) It is also worth noting that the boundaries of segments may become less distinct at higher angles of rotation (θ) of segments (defined in subsection 4.3.2), as explained in Figure 10 of the manuscript. This effect could potentially obscure the visibility of segment boundaries in electron microscopy making them hard to recognize in several species.
(2) In subsections 2.2.1 and 2.2.2, we have calculated the total number of layers and segments from the SEM image of N. globulus. Both the number of layers and the number of segments were determined to be 12. Afterwards, this value is retained for the skeletal reconstruction presented in section 4.3, depending on the chosen model (i.e., the layer model or the segment model).
(3) The term “twin lamellae” is used here to describe the repeating pair of lamellae (i.e., lamella-A and lamella-B) that together construct the full skeleton of Nannoconus. This term does not refer to crystallographic twinning, but rather to the microstructural arrangement and alternating inclinations of the lamellae. We acknowledge the ambiguity in this terminology and will provide a clearer explanation in the revised version.
In conclusion, we acknowledge that PXCT alone is not sufficient to definitively distinguish between the two models. However, it indeed supports the successful 3D reconstruction of the Nannoconus skeletal structure. When combined with existing 2D SEM observation, as well as biomineralization hypotheses, the evidence supports the interpretation of the Nannoconus's skeleton as being composed of “segments”, in the same way as Braarudosphaera is and other Mesozoic nannofossils (Aubry 2025 and forthcoming). We will clarify this point in the revised manuscript, along with an improved presentation of the geometric parameters and clear description of the ambiguous terms.
Rajkumar Chowdhury,
On behalf of all the co-authors,
Reference cited:
Aubry, M.P.: Cenozoic Coccolithophores: Braarudosphaerales, Atlas of Micropaleontology series Micropaleontology Press, 540 New York (336 pp), 2013.
Aubry, M.P.: Biomineralization in the Calcareous Nannoplankton Phenotypic Expressions Across Life Cycles, Geometric Control on Diversification, and Origin, Minerals, 15, 322, 2025.
Aubry, M.P.: Micalithophores: Braarudosphaerales, SEPM Concepts in Sedimentology and Paleontology, CSP XX, forthcoming, 2026.
Bralower, T. J., Monechi, S., and Thierstein, H. R.: Calcareous nannofossil zonation of the Jurassic-Cretaceous boundary interval and correlation with the geomagnetic polarity timescale, Marine Micropaleontology, 14, 153–235, 1989.
Hagino, K., Tomioka, N., Young, J. R., Takano, Y., Onuma, R., and Horiguchi, T.: Extracellular calcification of Braarudosphaera bigelowii deduced from electron microscopic observations of cell surface structure and elemental composition of pentaliths, Marine Micropaleontology, 125, 85–94, https://doi.org/10.1016/j.marmicro.2016.04.002, 2016.
Lees, J. A. and Bown, P. R.: New and intriguing calcareous nannofossils from the Turonian (Upper Cretaceous) of Tanzania, Journal of Nannoplankton Research, 36, 83–95, 2016.
Reinhardt, P.: Zur Taxionomie und Biostratigraphie des fossilen Nannoplanktons aus dem Malm, der Kreide und dem Alttertiär Mitteleuropas: mit 1 Tabelle, PhD Thesis, Dt. Verlag für Grundstoffindustrie, 1966.
Reznikov, N., Buss, D. J., Provencher, B., McKee, M. D., and Piché, N.: Deep learning for 3D imaging and image analysis in biomineralization research, Journal of Structural Biology, 212, 107598, 2020.
Van Niel, B. E.: Early CretaceousNannoconus’(calcareous nannofossil, incertae sedis) in NW Europe, University of London, University College London (United Kingdom), 1993.
-
AC2: 'Reply on RC2', Rajkumar Chowdhury, 25 Aug 2025
Peer review completion
Journal article(s) based on this preprint
Model code and software
Code (Python script) for the 3D submicron-scale skeletal reconstruction of Nannoconus Rajkumar Chowdhury, Alejandro Fernandez-Martinez, and Fabienne Giraud https://doi.org/10.5281/zenodo.14925063
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 894 | 173 | 45 | 1,112 | 33 | 42 |
- HTML: 894
- PDF: 173
- XML: 45
- Total: 1,112
- BibTeX: 33
- EndNote: 42
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Rajkumar Chowdhury
Redhouane Boudjehem
Baptiste Suchéras-Marx
Maxime Dupraz
Anico Kulow
Julio Cesar da Silva
Jean Louis Hazemann
Marie-Pierre Aubry
Javier Pérez
Alejandro Fernandez-Martinez
Fabienne Giraud
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2699 KB) - Metadata XML
The study “The 3D submicron-scale skeletal reconstruction of Nannoconus (Cretaceous calcareous nannofossil) - Insights on biomineralization” by Chowdhury et al. reports on the 3D imaging of Nannoconus shell, its segmentation into building units, and their computational reconstruction into a model shell. The work is very interesting as it gives the first 3D view into the abundant Nannoconus morphology observed in sediments. This brings an important addition to other structures of nanoplankton (most studied are probably the coccolithophorids). I very much appreciate the effort to use the segment morphology of the basic unit and to re-build from it in silico the full shell. The limitations of the study emerge from the fact that there is no living reference for the shells, and several assumptions are being made along the way. Nevertheless, this is a beautiful study that deserves publication after the following comments are addressed.
Comments: