A hundred years of dinosaur research in Mexico: skeletal completeness and phylogenetic information of the dinosaur fossil record in Mexico

1 Introduction

The dinosaur fossil record from Mexico is a good example of the gap between what Allmon et al. (2018) defined as “two fossil records”: an abstracted fossil record represented by the digital archives and databases, and a physical one comprising the specimens stored in museum collections. The abstracted dinosaur fossil record from Mexico is represented by 14 endemic dinosaur species and few referrals from other regions of Laramidia, whereas the physical dinosaur fossil record is very well represented in museum collections in Mexico (Ramírez-Velasco et al., 2014, 358). The abstracted fossil record, with low alpha-diversity, has been explained as a consequence of either the relatively low number of excavated Mesozoic outcrops of continental origin in contrast to the large number of trackways reported in the southern portion of Mexico (Rivera-Sylva et al., 2006, 240; Rivera-Sylva and Carpenter, 2014a, 152), or the low number of experts working on describing dinosaur fossils, the relatively recent interest in dinosaur paleobiology (Rivera-Sylva et al., 2006, 240), and, currently, attributed to sociological factors such as the difficulty of carrying out fieldwork due to the presence of organized crime. There has been also a set of paleobiological explanations to describe the low alpha-diversity of dinosaurs from the southern portion of Laramidia, such as a reduced number of paleogeographical regions (Rivera-Sylva et al., 2006, 237), or a paleoecosystem predominantly dominated by hadrosaurs (Rivera-Sylva and Carpenter 2014b, 176). These explanations follow the current line of research that focuses on large-scale questions by taking into account occurrences (e.g. Butler et al., 2009; Dunne et al., 2021; Griffin et al., 2022). However, occurrences have a lot of information that is not coded into the abstracted fossil record that reduce local signals —such as tectonic information, sedimentation regimes, hydraulic properties, but also research expertise and taxonomic effort — as well as creating artefacts (e.g. Kiessling, 2005; Noto, 2011; Close et al., 2020). When two dinosaur specimens with different completeness are identified, say to the species level, their actual physical state is obscured by the name in the abstracted fossil record, e.g. a maxilla and a partially complete skeleton become a line with equal weight (representing occurrence) in a database. Data wrangling of the abstracted fossil record (occurrences) before using does not necessarily rely on the physical fossil record, but on contextual data that is published elsewhere and reported with different degrees of details, thus, assuming research bias.

The production of identification guides, atlases, monographs and compilations creates a false sense of taxonomy as an end-user discipline that merely describes empirical data and packages it for its rapid use, as argued by Sluys (2013), instead of thinking of it as a constant revision of hypotheses and interpretations. The abstracted fossil record changes as the physical fossil record gets reinterpreted (Allmon et al., 2018). The phylogenetic information is derived from the taxonomic products but looks to model hypothetical common ancestor relationships, filtering out some taxonomic information. At its core, this is a database management problem where there has not been a way to convey the condition and preservation of the specimens associated with a taxonomic name to allow the user to discern between high- and low-quality specimens, or test the certainty associated with any taxonomic identification. This can be solved by adding more fields to a database, which requires standardization for them to become shareable, readable and comparable. However, another way to approach this is by developing a function of the data that shows the distribution of the information that could be extracted (phylogenetic information) from the physical fossil record. In a way, such function (an estimator) can help dimension not only the physical record, but also the research effort and interest put in the abstracted fossil record.

The global physical fossil record of dinosaurs comprises largely fragmentary and isolated remains as part of bonebeds, much more common than articulated skeletons. Eberth et al. (2007) were the first ones to examine bonebeds as a general phenomenon using large datasets in an exercise known as the Evolution of Terrestrial Ecosystems (ETE) Bonebed Database. Analysis of taphonomic information collated for the ETE Bonebed Database across 216 localities found no significant difference in the proportion of articulated remains since the Devonian, and the data does not support the hypothesis that the rise of Cenozoic bone-crunching predators increased carcass processing relative to the Mesozoic. The analysis concluded that, while bonebeds occur across a wide spectrum of terrestrial and marine depositional settings, the ETE Bonebed Database shows a pronounced dominance of fluvial environments, accounting for approximately half of all recorded bonebeds (often 60%–70% post-Carboniferous). This conclusion has also been applied to the Mexican dinosaur fossil record, suggesting that the southern portion of Laramidia represented a flood plain environment (Rivera-Sylva and Carpenter 2014b). Between the death of an organism, the decay of its soft- and hard tissues, the disarticulation, and the transport— as the remains become bioclasts suspended in the alluvium— taxonomically relevant characters get lost, affecting the final interpretation of the vertebrate fossils (Sansom et al., 2010, 2013; Sansom and Wills, 2013). Over the last two decades, several attempts have been made to qualify the completeness of the fossil record. The ‘Skeletal Completeness Metric’ (SCM) was developed as an attempt to map the completeness of sauropodomorph specimens through the Mesozoic (Mannion and Upchurch, 2010) and has since been employed in non-avian theropods (Cashmore and Butler, 2019), ichthyosaurs (Cleary et al., 2015), parareptiles (Verrière et al., 2016), and acanthodian fish (Schnetz et al., 2022). Although several implementations of the SCM have been developed, they are estimators; it is possible to make more accurate metrics by considering the volume of each individual bone, as has been done for human skeletons in forensics (Rowbotham et al., 2017). In this study, a new estimator is developed to capture the potential character richness of a bioclast based on the existing research, or its phylogenetic informativeness, to bridge the gap between the abstracted and the physical fossil records. The estimator developed here, known as phylogenetic estimator (pê), combines the skeletal completeness metric with a modified version of another metric, the character coverage metric (CCM) formulated by Mannion and Upchurch (2010, 285) as “the average percentage of characters attributed to each region of the body in (…) phylogenies”.

In this study, several statistical experiments were conducted to test the sensitivity of the phylogenetic estimator (pê) to different ways of computing the character coverage. A final statistical experiment was carried out to assess if the estimator pê approximates the fluvial depositional sub-environments. First, a separate dataset of Archosaurs from all the Mesozoic was downloaded from the Paleobiological Database (PBDB, https://paleobiodb.org) and the lowest taxonomic rank identified for each entry was paired with the type of fluvial environment as reported in the PBDB, to evaluate the loss of taxonomic information. The results from this experiment were then compared to the values computed for the different versions of the phylogenetic estimator pê for the Mexican fossil record to see if and how the estimator approximates the loss of taxonomic information.

1.1 Institutional abbreviations

BENC, Benemérita Escuela Normal de Coahuila, Coahuila, Mexico

CIC/P/, Colección Paleontológica del Centro del Instituto Nacional de Antropología e Historia en Coahuila, Coahuila, Mexico

CPC, Colección Paleontológica de Coahuila, Coahuila, Mexico

ERNO, Estación Regional del Noroeste de Sonora de la Universidad Nacional Autónoma de México

IGM-MG unofficial abbreviation, Museo del Instituto de Geología de la Universidad Nacional Autónoma de México, Mexico City, Mexico

IGM, Colección Nacional de Paleontología del Instituto de Geología de la Universidad Nacional Autónoma de México, Mexico City, Mexico

LACM, Los Angeles County Museum of Natural History, California, USA

MDCH, Museo del Desierto Chihuahuense, Chihuahua, Mexico

MUZ, Museo de Múzquiz, Coahuila, Mexico

SEPCP, Coordinación Paleontológica de la Secretaría de Educación Pública de Coahuila, Coahuila, Mexico

PASAC, Asociación Paleontológica de Sabinas Coahuila, Coahuila, Mexico

REG615PF, Registry number of a private citizen’s collection (Claudio de León Dávila) assigned by the National Institute of Anthropology and History, Mexico

1.2 Pronunciation guide

Following the recent recommendation by Brandt (2025) and to facilitate reading, Table 1 compiles the local pronunciation of the formations and geological features used in this text that are unfamiliar in the English literature. The list excludes the names of states and large regions in Mexico that have a standardized equivalent in English, like Yucatan, Sonora, Chihuahua, or Tamaulipas.

Table 1. Pronunciation guide in the International Phonetic Alphabet (IPA) of unfamiliar toponyms for Mexican geological features.

1.3 Overview of the Mesozoic sedimentary record in Mexico

As a starting point, a literature review was performed to reconstruct a tectono-geological history of the Mexican territory that allowed selecting the relevant geological formations for this study. The detailed version of this geological framework with the references it is based on is presented in the Supplementary Material. The depositional environments in the Mexican Mesozoic are all influenced by the opening of the Gulf of Mexico (Padilla y Sánchez, 2016). During the Late Triassic, three large faults formed in the southern portion of Laurentia and produced red bed deposits along the western coast of Pangaea (López-Ramos 1981; Fiduk, 2016) (Figure 1). No dinosaur remains have been found from these ages even though the plant communities of the Santa Clara Formation are similar to other North American paleofloras (Kustatscher et al. 2018), suggesting dinosaurs may have not reached these paleolatitudes yet. By the Early Jurassic, the sea transgresses through the middle portion of the coast as the magmatic arc of the Baja California Peninsula seals with the rest of the continent and possibly at the same time as the Yucatan Block moved along the Motawa-Polochic Transform Fault (Padilla y Sánchez, 2016) (Figure 2). This is where the dinosaur fossil record begins for the Mexican territory (Fastovsky et al. 1995; Munter and Clark 2006). By the Middle Jurassic, a rift valley is transgressed by the Panthalassa Sea and the Gulf of Mexico starts to open up until the rift became inactive by the Late Jurassic and a shallow sea formed separating the northern portion of the territory from the southern one (López-Ramos 1981; Fiduk, 2016; Padilla y Sánchez, 2016; Norton et al. 2016; Fillon 2016). A complex system of reefs and carbonated platforms formed from the Barremian to the Cenomanian, producing islands and archipelagos (López-Ramos 1981; Carrasco-Ramírez et al., 2016) that were inhabited by several types of dinosaur faunas (Figure 1). By the Turonian-Santonian, the dinosaur communities are more isolated as the sea transgression floods more of the islands and archipelagos and a deeper sea developed in the southern coasts of Laramidia (Murray et al., 1960; López-Ramos 1981; Porras-Múzquiz et al., 2014) (Figure 2). Terrestrial environments are then present throughout the country by the Campanian-Maastrichtian, as the Gulf of Mexico became disconnected from the Pacific Ocean. It is up to this point that the dinosaur faunas become better represented in the fossil record (López-Ramos 1981; Slattery et al., 2015; Padilla y Sánchez, 2016; Rivera-Sylva et al., 2016; Serrano-Brañas et al., 2016) (Figure 2).

2 Materials and methods

2.1 Data compilation

The record of dinosaur specimens was aggregated from those publications that reported Mexican fossils accessioned into a collection and identified to either as Ornithischia, Sauropodomorpha or Theropoda, including Mesozoic birds. Reports that suggested, for instance, “fragments of dinosaur bones” or “indeterminate dinosaur remains” were excluded. Contentiously described material was not excluded, for instance, the vertebrae found in the Campanian of Chihuahua initially described as titanosaurid by Montellano-Ballesteros (2003) also resemble hadrosaurid vertebrae, but this is not conclusive (D’Emic et al., 2010). The analysis only considers osteological remains, and as such, egg fragments and footprints were excluded from the aggregation.

Google Scholar (https://scholar.google.com) was primarily used for the search because, although it is independent from academic publishers and other metadata aggregators like Scopus or CrossRef, relying heavily on user curation, it is commonly used by academics due to its intuitiveness and readily access (see Smith et al., 2023). Clarivate and Web of Science require creating an account and several journals routinely used in paleontology, such as Revue de Paléobiologie, are not consistently indexed there. A second source of information was the Paleobiology Database (https://paleobiodb.org), particularly due to its emphasis on temporal and spatial information that is validated by editors. Lastly, a third source of information was the Léxico Estratigráfico Mexicano (LEM: https://sgm.gob.mx/Lexico_Es/), a list of all the lithostratigraphic units in Mexico on a government-supported free access website (Arriaga, 2025). In the LEM, each geological unit has an information sheet as downloadable PDF and includes both fossiliferous and non-fossiliferous outcrops. A survey through the available files allowed finding reports of dinosaur material documented in passing in other publications. Only the material that had an identification as a specimen, even if uncatalogued, was aggregated. The searches were performed between January and March 2026.

For Google Scholar, the search included the terms “Mexico dinosaur diversity”, “Mexico Mexican dinosaur dinosaurs” relying on publications that would include an English abstract, and “dinosaurios mexicanos” to fetch results in Spanish. The results from the search in English mostly led to uploaded documents in ResearchGate (researchgate.net), and publishers like mdpi.com and other aggregators like sciencedirect.com, scielo.com and wiley.com. Although MDPI has faced recent scrutiny over its publishing practices (see Abalkina, 2023), it is becoming commonplace in Mexican paleontology, thus the entries were not discarded by its provenance. The search in Spanish led to aggregators like redalyc.org and several science communication pieces published in university journals. In PBDB, the filters were set to “Triassic”, “Jurassic”, “Cretaceous” and “Dinosauria”. The LEM has a search engine that filters out by age and fetches exact matches. For instance, typing “Mesozoico” leads to only two units, namely Santa Eulalia (Baja California) and Rancho Viejo (Durango), whose age has not been constrained further into the Mesozoic. The word “Triásico” led to 31 results, “Jurásico” to 136, and “Cretácico” to 322. Units with continental influence were selected by navigating the LEM using the geological framework for the Mesozoic evolution of the Gulf of Mexico (Section 2.2; Figures 1, 2). This selection was complemented by referencing the "Columna Estratigráfica" feature in the LEM, an index covering all mapped units across the different states. Complete details of the geological framework are available in the Supplementary Material.

To include a study, a collection number or specimen number needed to be clearly stated. A couple of reviews (Ramírez-Velasco and Hernández-Rivera, 2015; Ramírez Velasco, 2021) have compiled a catalogue with collection data for dinosaur fossils reported in other sources without accession information, and the data reported therein is aggregated instead. Conference abstracts were excluded even if they mentioned the specimen number. Several studies were retrieved through ResearchGate since they were behind a paywall or otherwise access-controlled and were excluded if the available file was not a final print copy. Some papers were retrieved through shadow libraries. This process was done manually, and no automated tools were developed to find the papers. Taxonomic and collection information tends to be incorporated as part of larger studies and is often neglected in academic literature due to publication trends, so each paper retrieved was individually assessed looking for collection numbers and a description of the nature of remains, even if brief and vague.

Along with the specimen number or collection voucher, the following data was extracted from the studies: taxonomic status (if it is a type specimen, if it represents integumentary impressions, if the identification was done through open nomenclature, if there is an informal name given to the material), taxonomic determination — from more inclusive (namely Theropoda, Ornithischia, Sauropodomorpha) to species (going through family, subfamily and tribe) —, repetition, and the locality it was collected from (if mentioned, but if not, at least “Formation” needed to be available). The age of the locality was obtained from the most recent sources available, using geochronological methods, i.e. radiometry, magnetostratigraphy or biostratigraphy. The age reported in the publication or review was included in the database only if the locality was mentioned without a known Formation, or if the Formation had not been geochronologically dated. This was also not automated since this data is non-standardly reported. Specimens in collections not reported elsewhere were only incorporated if the authors confirmed they saw the specimens first-hand. The list of sources analyzed are in the Supplementary Material.

2.2 Skeletal completeness metric

The skeletal completeness is an estimator to report how much a skeleton is preserved and is used both in extant and fossil vertebrate remains. One approach, like the one developed for human forensics, relies on using average volume of human bones and assign each bone a rough percentage (Rowbotham et al., 2017, tbl. 1), with the advantage of having a better estimator for repetitive units with different sizes, for instance, metacarpals 2 (both elements averaging 5.6 cm³) and 3 (both totaling 8.8 cm³) getting each a skeletal completeness metric (SCM) of 3%, whereas metacarpal 4 (5.8 cm³) has a SCM = 0.2%, and metacarpals 1 (5.6 cm³) and 5 (5.4 cm³) both get a SCM = 0.18%. This granularity in an estimator is easier to do when having the same species, but becomes convoluted when dealing with dinosaurs, where the different clades have drastically different volumetric proportions between the skull and the bodies. For Sauropodomorpha, Mannion and Upchurch (2010, tbl. 1 and Supp.) assigned skeletal completeness metrics of 10% for the skull, 55% for the axial skeleton, and 35% for the appendicular skeleton. The skeletal volume can vary greatly between species, as allometric patterns can change through ontogeny, and across clades: for instance, whereas sauropod skulls can be very reduced and would represent less than 10% of the skeletal volume, in ornithischians like Triceratops, it can represent up to 33% of the skeletal volume. As a database management technique, this granularity would require a forced recomputation if, for instance, a titanosaurid (e.g. Montellano-Ballesteros, 2003) is reinterpreted as a hadrosaur (e.g. D’Emic et al., 2010).

Therefore, the skeletal completeness metric (SCM) was standardized to be applicable across the different dinosaur clades despite the widely variable skull, axial skeleton, and limbs proportions. The total score of 100 is distributed along the different regions of the body: a score of 10 is assigned to the skull, 51 to the axial skeleton, 35 to the appendicular skeleton, and 4 to non-skeletal bone, such as osteoderms, ossified tendons and ossified cartilage. The skeletal completeness metric (SCM) in the skull is distributed along the cranial regions: the upper arcade lateral series (SCM = 1, comprising premaxillae and maxillae), the circumorbital series (SCM = 2, lacrimal, prefrontal, postfrontal, jugal and postorbital), the median series (SCM = 2, nasal, frontal, parietal, post-parietal and the bones of the palate), the cheek series (SCM = 1, squamosal, quadrate, and quadratojugal), the braincase (SCM = 0.5), the mandibular series (SCM = 2.5, predentary, dentary, angular, surangular, articular, and splenial), and the teeth (SCM = 1). The skeletal completeness of the axial skeleton is then subdivided into four regions (cervical vertebrae, SCM = 15; dorsal vertebrae, SCM = 15; sacral vertebrae, SCM = 5; and caudal vertebrae, SCM = 16), each considering vertebral bodies and ribs or chevrons as subdivisions. The scores for the appendicular skeleton are distributed along the pectoral girdle (SCM = 5), forelimbs (SCM = 12), pelvic girdle (SCM = 6), and hindlimb (SCM = 12).

When compiling the data, the skeletal completeness metric was assigned according to what was reported accounting penalization for “incomplete” bones, and for the number of bones. For instance, two complete femora account for a SCM = 4; if a paper reports one distal half of a femur, then it is scored as SCM = 1. In some detailed descriptions, e.g., the monographic description of a new species, the assessment is easy, but when a paper mentions in passing the completeness of a bone, this is done to best capture the nature of specimen, even if overestimated, for instance, “limb remains” are scored as SCM = 24 for both fore and hindlimbs. However, in other cases, such as compilations or lists in reviews, the description is given as “fragmentary femur” or “large bones”; in these cases, for instance, the penalization gives values like “0.5” to capture the presence of one femur.

2.3 Character coverage metric

Here, the character coverage metric (CCM) is modified to account for the total number of elements needed to assess the hypothesis of homology formulated by each character statement. If a character statement establishes the assessment of the connection between two bones, then that character statement is given a higher weight than one requiring only one bone to assess the hypothesis of homology. For example, the character list by Baron et al. (2017) with 456 character statements has a character coverage of 531, meaning that some characters require multiple elements to be confidently evaluated. Some characters refer to overall regions like “skull” or “supratemporal fenestra”. The tally is then categorized into the same regions defined for the skeletal completeness metric (SCM), allowing for character statements to be represented in several regions. If a character requires assessing the maxilla and the jugal, then the tally gives “1” to each the lateral bearing tooth series and the cheek series; the rationale of this flexibility is that, for example, anatomist may be able to infer the shape of the jugal when absent based on the morphology of the maxilla.

Because the skeletal completeness metric has been optimized, e.g. a complete skull always scored as 10 regardless of the actual volumetric proportions, the character coverage metric should complement this to give an idea of the phylogenetic character richness of any specimen. For example, in the character list by Dieudonné et al. (2020) for ornithischians, the lateral tooth-bearing series (rostral, premaxilla, maxilla) and the mandibular series (predentary, dentary, angular, surangular, articular and splenial) have 29 characters each, the cheek series (squamosal, quadrate and quadratojugal) has 33 and the circumorbital series (lateral, prefrontal, postfrontal, jugal and postorbital) has 37, meaning that five bones in a skeleton (circumorbital series) are required to assess 37 hypotheses of homology, whereas only 3 bones in the cheek series are needed to assess a similar number. Each region was defined a priori for all the skeletons (see Supplementary Material) and for each character list, the primary locators and qualifiers (sensu Sereno 2007) were extracted and counted, so that some characters are applied to entire regions (e.g. skull, orbit, infra- and supratemporal fenestra) whereas others are applied to several bones. For Dieudonné et al. (2020), 286 characters need 342 evaluations on the skeletal elements (Supplementary Material, Table 2).

Initially, it seems that this is a very dynamic metric, with each character list having a different coverage, but within clades, the number of characters differs little. For instance, phylogeneitc matrices for sauropodomorphs often differ in their taxon sampling, have a few added characters, and feature modified scorings, with the character list expanding incrementally across iterations (Regalado Fernández and Werneburg, 2022). A statistical study on phylogenetic matrices to include all dinosaurs found that the matrices do not have statistically significant differences in the models they produce (Černý and Simonoff, 2023), suggesting that within a clade, the character coverage would not differ drastically. However, the character coverage will be different at different hierarchical levels, for instance, between sauropodomorphs and ornithischians and theropods, because hypotheses of homology are differentially distributed for each group.

For this study, only character lists that were published as either part of the paper, an appendix or as associated Supplementary Material were analyzed. It is quite common that papers with phylogenetic analysis only show the final matrix with the taxa used but no character statements are appended to the file (Godoy et al., 2026, Figure 3). For the selected character lists, each character statement was decomposed in the osteological elements needed to assess it and the counts were categorized using the same regions defined for the SCM (Supplementary Material, Tables 1-9). The matrices chosen have the following hierarchical focuses: Dinosauria (Baron et al., 2017), Ornithischia (Dieudonné et al., 2020), Theropoda (Rauhut, 2003), Sauropodomorpha (Pol et al., 2021), Ornithopoda (Madzia et al., 2020), Marginocephalia (Butler et al., 2011), Thyreophora (Sánchez-Fenollosa and Cobos, 2025), and Titanosauria (Gorscak and O’Connor, 2019).

2.4 Phylogenetic estimator

A phylogenetic estimator should be a function that differentiates between specimens, estimates the usefulness of the occurrence as a datapoint and filters them according to the assumptions needed for analysis. The phylogenetic estimator, pê, should reflect the spread of completeness and the phylogenetic character richness. Here, this estimator pê is computed as a cross product of the skeletal completeness metric and the character coverage (SCM × CCM). For instance, having two ilia from the same individual is more informative to score a character than only one ilium, because it is possible to differentiate preservation or deformation by evaluating the character twice, so the proportional value of skeletal completeness for one ilium is 3.45%, multiplied by the character coverage of the respective character list. To test the robustness and sensitivity of the estimator, four experiments were conducted to see how it is affected by the character list and its phylogenetic hierarchy scope and what the value of the estimator communicates using the compiled dinosaur fossil record from Mexico. The phylogenetic estimator then was used to test the intuition that the more incomplete the material is, the less phylogenetically relevant information it yields.

The first experiment calculates the phylogenetic estimator pê as a constant applied to all values of skeletal completeness, using the character list by Baron et al. (2017) that compiled characters pertaining to the three lineages of dinosaurs. To plot the correlation between the estimator and the skeletal completeness metric, the values were ranked through a minimum-maximum normalization (pê* to differentiate from non-standardized pê), so that the most complete specimen in the dataset gets a value of 100. The second experiment considers the character coverage of the Dinosauria-focused dataset plus the character coverage of three matrices with a scope in Ornithischia (Dieudonné et al., 2020), Theropoda (Rauhut, 2003) and Sauropodomorpha (Pol et al., 2021), and the normalized pê* is calculated for the new values. The third experiment considers what happens when characters are restricted to a single intra-Dinosauria lineage. And the fourth experiment evaluates when a mixed-scope is approached, by calculating the character coverage from the character lists for Ornithopoda (Madzia et al., 2020), Marginocephalia (Butler et al., 2011), Thyreophora (Sánchez-Fenollosa and Cobos, 2025), Neotheropoda (Dalman et al., 2024) and Titanosauria (Gorscak and O’Connor, 2019), as well as the three higher-rank character lists from Ornithischia (Dieudonné et al., 2020), Theropoda (Rauhut, 2003) and Sauropodomorpha (Pol et al., 2021) for partially determined specimens.

Finally, two experiments were conducted to test the usefulness of the phylogenetic estimator pê. In one experiment, the normalized values of the estimator (pê*) were plotted against time to compare their trends with those of the skeletal character completeness. To visualize the correlation of SCM and the estimator pê over time, these values were divided into time bins of 4.9 My from 184 Mya (Toarcian, Early Jurassic) to 66 Mya (Maastrichtian, Late Cretaceous). For each specimen, the locality was associated with a geological formation. The age for the formation was taken from the most updated estimation performed, even if it did not match the age suggested in the original paper (see Supplementary Material). Then, the tally counted the number of specimens that would be included in each time bin based on the category they fell in, for example, “late Maastrichtian” specimens tallied for the time-bin 73 to 66 Mya.

A fifth experiment was performed to see if there were any similarities between the depositional environments with the lowest skeletal completeness metrics and phylogenetic estimator in the Mexican outcrops was consistent with a global Mesozoic-wide database of archosaurs in terms of assigned taxonomic hierarchy. All occurrences of Archosauria (fossils, ootaxa, and ichnites) for the time span between the Ladinian (240 Mya) and the Late Maastrichtian (66 Mya) were selected. From the 27,070 records, 23,936 (around 88%) are registered in collections from terrestrial environments, of which 8,166 (nearly 60% of the occurrences in terrestrial environments) are associated with fluvial deposits. The number is likely higher, since most records (11,745 entries) are classified as “indeterminate terrestrial environments” or simply “terrestrial” (full data set as Supplementary Material).

3 Results

3.1 Database of Mexican fossil record

The caveat of the method compilation to aggregate data is that several biases in the literature are likely to be amplified. For example, the final tally of specimens aggregated was 673, and over half of these entries are identified as Hadrosauromorpha (n=365). One explanation for this bias is that most of the dinosaur fossils come from the Late Cretaceous, when the Mexican territory coalesced from being an archipelago into a long peninsula extending from the south of Laramidia (Figure 2). Hadrosaurs originated around the Coniacian and became the last large diversification event of dinosaurs reaching its peak by the Campanian (Lund and Gates, 2006). In Laramidian outcrops from this age, most of the hadrosaur remains are found disarticulated in bonebeds, so their accession number can be inflated by having each remain accessioned as one specimen or underestimated by giving a collection number to the whole assemblage. Moreover, the commonality of the remains is increased by the regional taxonomic expertise, which is more developed towards hadrosaurs than to other dinosaurs (Ramírez Velasco, 2021, Figure 3). By association, it is also possible that non-hadrosaur remains are accessioned as hadrosaur pending further research into their identification. Several entries are reported as in situ even if they have been formally accessioned into a collection, meaning that their report as “hadrosaur” can be seen as a preliminary. This taxonomic expertise bias could also be confounded by preservation in the absence of other groups. For example, in the Huizachal Canyon, Tamaulipas, there are reports of heterodontosaurids (Clark et al., 1994) and a coelophysoid (Munter and Clark, 2006) that have been given collection numbers, but there are also reports of large dinosaur bones that, given the age and size, may correspond to early-diverging sauropodomorphs (Munter and Clark, 2006, 53–54). Similarly, for the Otlaltepec Formation, there are reports of 81 fossil fragments, but only three of them have been described as one specimen and identified as a diplodocoid remains (Rivera-Sylva and Espinosa-Arrubarrena, 2020). These 78 potentially sauropodomorph entries are excluded (no collection number and no identification to any of the dinosaur clades), leaving the dataset with only 6 specimens identified as Sauropodomorpha. The absence of taxonomic identification can be confounded by actual sampling and preservation biases that have been explained through biogeographic and geological explanations, such as the “Laramidian sauropod hiatus” (Mannion and Upchurch, 2011). Thus, the low occurrence of fossils in one region translates through the supervisor-student relationships (Ramírez Velasco, 2021, Figure 3) into lower taxonomic expertise that amplifies the low identification of sauropodomorphs.

3.2 Skeletal completeness metric

The distribution of the skeletal completeness metric (SCM) of osteological remains from Mexican dinosaur fossils has an average of 1.71 ± 0.162 through the Mesozoic; however, with a median of 0.6 and with 75% of the datapoints having skeletal completeness lower than 1, the data is not normally distributed with outliers having a strong pull upward. Nevertheless, skeletal remains identified to a species level (Table 2) are not restricted to outlier positions and represent 11.9% of the total, of which only 14 specimens are treated as type material. The most complete holotype is that of Velafrons coahuilensis, with a completeness score of 43.25, the second-highest outlier, whereas Xenovenator espinosai (SCM = 0.5) – a braincase from a troodontid (Sullivan, 2006; Aguillón-Martínez and Rivera-Sylva, 2023; Rivera-Sylva et al., 2026) – has a more common completeness higher than 25% of the specimens. The most complete specimen recorded so far in Mexico corresponds to specimen PASAC-1, informally known as “Sabinosaurus” or “Sabinasaura” (SCM = 49.6), and although it still has not been formally described, it is treated as a separate species in the literature. Another informally described specimen is “Saltillomimus rapidus” (SCM = 15.64), although the name was given in a dissertation and has not been formally described (Table 3).

Table 2. Character coverage for each of the character lists selected for this study, tallying the number of times each region is needed to assess a hypothesis of homology (character statement).

(A)Dinosauria in Baron et al. (2017), (B) Theropoda in Rauhut (2003), (C) Neotheropoda in Dalman et al. (2024), (D) Ornithischia in Dieudonné et al. (2020), (E) Ornithopoda in Madzia et al. (2020), (F) Marginocephalia in Butler et al. (2011), (G) Thyreophora in Sánchez-Fenollosa and Cobos (2025), (H) Sauropodomorpha in Pol et al. (2021), and (I) Titanosauria in by Gorscak and O’Connor (2019).

Table 3. Skeletal completeness metric (SCM) of the specimens that have been treated as the type specimen and hypodigm for names reported only in the Mexican territory.

3.3 Character coverage metric

Isolated remains can be useful if they enable determining the diversity of an assemblage, however, they may not provide a phylogenetic signal if, for instance, none of the character lists used those elements. The atlas and axis represent a skeletal completeness of 2, but if it is identified as a neotheropod, the character richness is as high as that of the postaxial cervical vertebrae (Table 2), but if it is identified as a thyreophoran, it is not used as a source of homology testing in the lists employed here. This essentially means fragmentary material is noisier since the phylogenetic information they convey will depend on the taphonomic processes that produced them (disarticulation and association to the rest of the skeleton, commonality of the preservation, retention of diagnostic features). When specimens from the Mexican dinosaur fossil record with a skeletal completeness metric lower than 20 are removed from the sample, the correlation between the phylogenetic estimator (pê) and the skeletal completeness becomes noisier (Section 4.4), but their inclusion tightens the correlation, suggesting that skeletal completeness generates high-leverage points that corroborate the intuition that more completeness yields more phylogenetic information. This effect may occur with other metrics when applied to large assemblages or large bonebeds.

The level of identification of fragmentary remains that come from body regions where fewer characters are discovered varies greatly in the Mexican dinosaur fossil record. The most complete skull is that of Tlatolophus (SCM=8.8), the skull being the richest area for the Ornithopoda character list (Table 2). Most of the braincase elements have been assigned to Ornithopoda, also an area rich in characters (Table 2). Most of the cervical vertebrae with taxonomic identification belong to Saurolophinae, followed by Hadrosauromorpha; cervical vertebrae are more phylogenetically informative at the broader Ornithischian level than for differentiating among specific ornithopods (Table 2). The caudal vertebrae are mostly identified for Ornithomimidae and Hadrosauridae, both with a high character coverage in Neotheropoda and Ornithischia respectively (Table 2); and of other tail elements, most of the chevrons are identified as Ornithischia. The pelvic girdle elements are also more likely to be identified as Hadrosauridae.

3.4 Phylogenetic estimator (pê)

Intuitively, practitioners sense that as the skeletal completeness increases, so will the phylogenetic informativeness. The correlation between the completeness metric and the normalized scaled estimator pê (pê*) was assessed through two sets of four experiments changing the way in which the pê was computed, one set with all the specimens included, and a second one removing all specimens with a completeness metric equal to or greater than 20.

The skeletal completeness metric has a strong positive correlation (r) when the estimator pê* is computed using the character lists compiled for Dinosauria (Supplementary Table 1) (pê_D, r=0.93), for Dinosauria plus either Ornithischia (Supplementary Table 2), Theropoda (Supplementary Table 3), or Sauropodomorpha (Supplementary Table 4) (pê_D+OTS, r=0.93), and for only Ornithischia, Theropoda and Sauropodomorpha (pê_OTS, r=0.91), explaining over 80% of the variance (r² = 0.8806, r² = 0.8807, r² = 0.8328, respectively, p = 0.0001). The correlation becomes weaker when using the character coverage computed from the lower hierarchical levels, namely Neotheropoda (Supplementary Table 5), Ornithopoda (Supplementary Table 6), Marginocephalia (Supplementary Table 7), Thyreophora (Supplementary Table 8) and Titanosauria (Supplementary Table 9) (pê_low, r=0.6403) explaining only 41% of the variance (r² = 0.40999, p=0.0001) (Supplementary Table 10). This indicates that fragmentary remains can either retain a strong phylogenetic signal or lack one entirely, even if more complete material is generally more informative. The same correlations were then evaluated using a reduced set, removing 12 specimens that had a skeletal completeness equal to or greater than 20 (Supplementary Table 11). The correlation was still positive when the estimator pê* is computed with the first three treatments (pê_D, pê_D+OTS, pê_OTS, with r=0.77, 0.77 and 0.71 respectively) but is significantly weaker (r=0.48) when using the character lists from lowest taxonomic rank pê_low) (Supplementary Table 11). The variance explained by the correlations also degenerated quickly with each treatment, going from 0.6 (with the character coverage for all Dinosauria) to 0.24 (with the character coverage from lowest taxonomic hierarchy). When visualizing the last experiment (Supplementary Figure 8), the specimens identified as theropods have a strong positive correlation with the phylogenetic estimator pê whereas the ornithischians have persistent low pê values. The latter is consistent with the disagreements reported in the literature on the taxonomic identity of hadrosaur remains.

Noisiness in the data can be appreciated by looking at the skeletal completeness and estimator pê in the specimens identified to a species level. Velafrons coahuilensis shows a progressive decline in the estimator (pê_D=93.8, pê_D+OTS=86.8, pê_OTS=73.1, pê_low=54.6) suggesting that despite its completeness (SCM = 43.25), the skeleton has fewer elements that can allow testing homology hypotheses for Ornithopoda relative to other superior taxonomic levels such as Dinosauria (highest) or Ornithischia (Table 3). The opposite is true for Labocania aguillonae and L. anomala; the former has a completeness score of 11.75, and the estimator pê drastically improves twofold when using lowest rank character coverage metrics (pê_low=64.1) in contrast to values around 30 for the higher ranks (Table 2*). A different trend is seen in Mexidracon longimanus (SCM = 25.05) whose pê* estimators decrease in character coverage computed for higher ranks but becomes the one with the highest phylogenetic character richness when using the character list for Neotheropoda (pê_low=100). The pattern becomes more chaotic when using fragmentary remains referred to other type species (Table 4*).

Table 4. Specimens that have been identified as belonging to species whose types were discovered outside Mexican territory.

The distribution function of the estimators has a similar shape than the skeletal completeness. The computation that applies the same character coverage values for all entries (pê_D) has an average of 4.348 ± 0.335, 75% of the entries have a pê_D ≤ 4.53, and a positive skewness caused by outliers; the most common value is 3.9, and has a similar kurtosis to SCM (50.58, 51.33 respectively), suggesting that the outliers are more frequent and extreme. The coefficient of variation (CV) is lower for pê_D* (CV = 200.05) than that of SCM (CV = 246.8), suggesting this version of the estimator is less volatile than SCM. The CV of the estimator computed using only middle rank phylogenetic characters (Ornithischia, Theropoda, Sauropodomorpha) (pê_OTS) is a bit more volatile than pê_D (CV = 202.43), whereas the estimator computed with the focus on the lowest taxonomic ranks (pê_low), with a CV = 271.92, is more volatile than all the other estimators and SCM. Only the estimator pê_low has a considerably higher kurtosis (62.312) than the other ones and SCM, suggesting that only at this level the outliers are more frequent and extreme (black swan events) (more details in Supplementary Table 12).

3.4.1 Phylogenetic estimator through the geological scale

Since the estimator with the lowest volatility is the one using the character lists for Dinosaurs plus Ornithischia, Theropoda and Sauropodomorpha, pê_D+OTS, this version was chosen to assess its behavior through time. The highest average pê_D+OTS in the dataset is assigned to Avialae (pê_D+OTS*=11.91, Tables 4 and 5), corresponding to two specimens dated for the Turonian. Sauropoda also has a high average (pê_D+OTS=10.18), the highest values corresponding to the diplodocoid metatarsals from the Bajocian-Callovian (pê_D+OTS=21.09), and fragmentary remains spanning from the late Campanian to the early Maastrichtian. Ornithopoda has a high average pê_D+OTS* of 4.21, but also has the widest range, going from 100 (meaning this is the highest phylogenetic character coverage [CCM] in all the samples) to 0 (meaning the material found did not map onto corresponding characters). The distribution of skeletal completeness is similar to that of pê_D+OTS* (Figure 3; Table 6).

Table 5. Summary statistics of the estimator for phylogenetic informativeness for all the clades identified in the Mexican dinosaur fossil record.

The oldest dinosaur remains in the Mexican fossil record are documented from La Boca Formation (Tamaulipas) — coinciding with the fault movement along the Motawa-Polochic Transform Fault and the first sea transgression in the Portal del Balsas, producing a floodplain (Figures 1, 2)— and although fragmentary, their phylogenetic relevance is quite high (Figure 3). The Middle Jurassic “Bajocian-Callovian” interval trend corresponds to an extrapolated value from the diplodocoid metatarsals from Otlaltepec. This “jump” is characteristic of the high values seen in the next time bins. A similar situation occurs in the Ocozocuautla Formation, where several remains have been reported (Carbot Chanona and Avendaño Gil, 2002), but only maniraptoran remains (average pê_D+OTS=0.81) have been formally described (Carbot Chanona and Rivera Sylva, 2011). The remains deposited in fluvial environments probably derived from continental drainage onto the shallow-water carbonate platform (Figures 1*, 3, 4).

Table 6. Summary statistics (avg = average, min = minimum, max = maximum) of the skeletal completeness metric at the specimen level (SCM) and the normalized scaled estimator for phylogenetic informativeness using the Dinosauria, Ornithischia, Theropoda and Sauropodomorpha character coverage metric (p̂D+OTS*) over time bins ranging from 184 Mya to 66 Mya.

3.4.2 Loss of taxonomic information in fluvial sub-environments

A proxy to measure taxonomic information in current databases can be how far down the taxonomic ranks the material is identified; it can be assumed that a diagnosis only to the level of Class may be related to either a very eroded fragment or lack of local expertise in the region, although given the PBDB is an aggregate of peer-reviewed publications curated by expert editors, the second bias is mitigated. From the terrestrial records of Archosauria obtained from the PBDB, 8,166 (≈34.1% of terrestrial, or ≈30.2% of total) are explicitly associated with fluvial deposits, a number likely underreported by 11,745 entries categorized as “indeterminate terrestrial environments” or “terrestrial.” Within the identified fluvial systems, most occurrences are recorded in floodplains (Figure 4), followed by fluvial-lacustrine, channel, and channel-fill settings. An additional ≈20% of the fluvial registries are categorized as “fluvial,” precluding precise sub-environmental assessment (e.g., distinguishing channel bottom from other sub-environments) (Figure 4). This distribution highlights the dominant role of fluvial systems in preserving Archosaurian fossils, while simultaneously underscoring significant taphonomic biases and data limitations in sub-environmental determination.

Of the 27,070 records downloaded, 7,990 are identified to species level (29.5%) and 4,581 (16.9%) are identified to the genus level. One hundred eighty three records are identifiable only to class level (Aves). While their gross morphology clearly indicates avian remains, they lack the diagnostic traits (or taxonomic expertise) required for finer identification. This analysis suggests a strong partitioning of preservation quality across sub-environments: the lowest taxonomic rank – Class – is most common in crevasse splay deposits and secondarily in channel deposits, indicative of highly fragmented or poorly preserved remains (Figure 5). In contrast, Species-level resolution is predominantly restricted to fluvial-lacustrine environments, where this preservation category is notably rare in other settings. Genus-only identifications are most frequently found in channels and floodplains (Figure 5). This distribution highlights how specific fluvial sub-environments act as distinct taphonomic filters of taxonomic information, with low-energy lacustrine settings significantly favoring the preservation required for high-resolution (Species-level) taxonomic identification.

In the case of the Mexican dinosaur fossil record, the skeletal completeness metric and the phylogenetic estimator pê follow a similar pattern in their relation to the energy levels and fluvial sub-environments. The largest average skeletal completeness metric comes from low-oxygen deltaic floodplains (6.0), followed by alluvial continental (4.2), and sequences that represent transgression sequences from continental shelf, to prodeltaic, to shoreline (3.7). The lowest average skeletal completeness metric comes from channel and point-bars identified as meandering rivers (0.6), fluvial deposits (0.4) and shoreline and inland environments (0.2). Several specimens with a large completeness score come from unspecified depositional environments (average SCM = 6.0). The specimens with a skeletal completeness equal to or larger than 20 come from fluvial environments: marshes and deltas (49.6), regression sequences from shoreline to alluvial floodplains (43.3, 32.0, 26.8, 25.1, 24.3, 22.5, 22.0, 20) sequences that change from braided to sinuous fluvial environments (28.3), and floodplains (23.3). Using the same version of the estimator pê (pê_D+OTS), the highest average phylogenetic estimator corresponds to fluvial deposits, marginal lagoons, open lagoons and deltaic (minimum 5.2, maximum 30.4), followed by fluvial-lacustrine floodplains (minimum 4, maximum 18.4), and alluvial continental (minimum 0, maximum 20.2). In the PBDB experiment, the wet floodplain has the highest proportion of identifications to the genus level (Figure 5), and the fluvial-lacustrine the highest proportion of species identifications. The lowest phylogenetic information estimators correspond to in continental shelves and prodeltaic (minimum 0, maximum 2.3), and shoreline environments (minimum 0, maximum 1.1). In the PBDB, the prodelta, delta front and interdistributary bays (Figure 5) have the lowest proportions for any identification. Loss of taxonomic information does not seem to be solely related to the energy level in an environment (Figure 5) but also with the particular local dynamics, such as sedimentary intake, sedimentation rate and hydric regime, but it does seem from these two experiments, that fluvial-lacustrine, fluvial, channel and crevasse plays have the highest probability of producing species-level identifications (Figure 5*). What these features have in common is the behavior of water over the topography, a stream that moves through the lowest points, albeit at different scales. These results are consistent with what was initially reported for the ETE Database Analysis performed by Eberth et al. (2007).

4 Discussion

The gap in the osteological record between the Late Jurassic and the Early Cretaceous (Figure 3) is consistent with the environments associated with the rifting processes during the opening of the Gulf of Mexico. Dinosaur trackways are recorded on the northern and southern continental margins of the Gulf of Mexico (Figure 1, localities 5, 6, 14). The preservation of trackways along with the absence of osteological material is similar to the conditions seen in the Triassic-Jurassic Connecticut Valley, USA, where the bones may have dissolved prior to lithification due to low pH conditions produced by the root transpiration and intense precipitation (Galton et al., 2025); the likelihood of fossilization reduces if there was an intense drainage system where any bone deposited into the shallow platforms experienced intense erosion or transport. Between the Barremian-Aptian the osteological record is so far inexistent (Figure 3). This corresponds with the end of the rifting in the Gulf of Mexico, and the formation of a complex system of islands, archipelagos and reefs that covered most of the territory (Figure 1, “Barremian-Aptian”). The dinosaur fossil record, both osteological and ichnological, reappears by the Albian, coinciding with the coalescence of islands and archipelagos, and the continental influence in the Gulf of Mexico Basin. Shallow waters (Figure 1, localities 3, 7, 10) suggest constant disconnection and reconnection of the landmasses enabling the dispersion of dinosaur groups, and some places like the islands surrounding the San Martín Atexcal paleo-reef may only preserve trackway records (Figure 1, locality 12). The osteological record shows the largest ranges of completeness during the Late Cretaceous (Figure 2), with more developed fluvial systems that enabled different types of preservation. Interestingly, this coincides with the lowest record of trackway localities (Figures 2, localities 22, 24) (Rodríguez-de la Rosa et al., 2004).

While preservation bias is increasingly recognized as a primary driver of the structure of the fossil record, determining whether a signal in the fossil record reflects an authentic biological phenomenon requires accounting for more than just taphonomic bias. Along with colonial practices that have shaped the priorities when collecting data, as it has been recently discussed elsewhere (e.g. Raja et al., 2022; Chacon-Baca et al., 2023; Monarrez et al., 2022), our interpretations are equally shaped by data management and the cumulative nature of database construction. Regional taxonomic expertise, for instance, significantly dictates which skeletal elements are recognized and prioritized for reporting, introducing a filter before the dataset building, followed by a publication filter that often results in the exclusion of unidentifiable bioclasts from the abstracted record. As demonstrated by the assemblages of Otlaltepec and Ocozocuautla Formations, when these materials are not recorded, the resulting absences in diversity, distribution, or dispersal patters are frequently misinterpreted as true stratigraphic gaps. When conducting large-scale data analysis, it is important to account for the initial filter created during the reporting of taxonomic names in large databases. This filter is a byproduct of taxonomic expertise and regional research history of what is known as taxonomic impediment (De Carvalho et al., 2007). The prevalence of hadrosaur remains for the Campanian-Maastrichtian for Mexico reported in large datasets like the PBDB may not strictly refer only to their original ecological abundance or the prevalence of their preferred environments but rather compounded with the long history of hadrosaur research in Mexico (Ramírez Velasco, 2021), which increases the documentation, identification and cataloguing of material that can be recognized as Hadrosauromorpha.

In the coastal lowlands, there is a significant trackway record of sauropods, theropods and ankylosaurs in the Tecomazúchil Formation and near Tlaxiaco (Zorrillo-Taberna Indeferenciadas Formation) (Lozano-Carmona et al., 2021; Rodríguez-de la Rosa et al., 2004), suggesting that the Otlaltepec Formation was a depositional environment from the drainage system in the coastal lowlands and the other localities have a low osteological record product of this hydric regime. Although fluvial environments are the most common setting to find vertebrate remains in the Mesozoic, rivers are inherently complex sedimentary systems characterized by significant spatiotemporal variation across depositional environments with varying levels of energy. In general, channel stability decreases as sediment supply and caliber increase (Church, 1992; Erskine, 1999). Although often reported together in the data collected for this study, meandering and braiding function as continuous alternating processes within the floodplain rather than representing discrete, linear stages of river evolution (Ferguson et al., 2022). Drainage in rivers is largely understood through a ‘sediment cascade’ model, in which as the drainage area increases, grain sizes decrease and the volume of alluvium increases (Church, 1992; Ferguson et al., 2022). Bones and skeletons become bioclasts once they become part of the taphonomically active zone, and as such, their shape will alter their hydrodynamic behavior in rivers or lakes. An assortment of similarly sized bones may correlate greatly with the drainage area, and the larger the alluvium transported, the more likely it is for certain bones to accumulate. In concert, vertebrate skeletons are also complex, composed of elements with a wide spectrum of sizes, shapes, densities, and the tendency to be transported or scavenged. It is therefore expected that different fluvial sub-environments preserve unique associations of vertebrate skeletal elements, with varying degrees of time-averaging, and thus record different components (and overall views) of the original ecosystem. Because bone shape dictates hydrodynamic behavior, preservation potential varies across the diverse sub-environments of the river system.

To mitigate these biases, it is important to incorporate more information associated with taxonomic names. The skeletal completeness metric is a good way to characterize the completeness of specimens into databases. When new species or first report of specimens are published, there is always an analytical section that describes the completeness of a specimen and the certainty of the taxonomic identification. In the PBDB, for instance, this information gets somewhat lost by having specimens described as “partial skeleton”. In large datasets, coding for this information allows for the identification of high-leverage outliers. A focus on the environments that produce extreme outliers, defined here as anything with a skeletal completeness equal to or larger than 20, risks distorting the broader environmental signal provided by more common elements with lower completeness values. Further studies on how skeletal completeness is related to these taphonomic factors would help to develop a better picture of skeletal assemblages, e.g. time averaging or provenance of the remains.

Furthermore, the estimator of phylogenetic informativeness (pê) could enable better decision making when cleaning datasets before use. For instance, if a bone is identified as Maniraptoran with a skeletal completeness of 2, but a phylogenetic estimator pê of 0, this datapoint can either be removed from the analysis, or used to evaluate the sensitivity of it or identified as a source for further research. When describing a bonebed, it may allow the evaluation of the potential depositional events and origins of the carcasses. For example, a dataset that contains the data points Edmontosaurus (n=30) and Troodon (n=2) gives an equal weight to each of the findings and may lead to the interpretation of it as a paucispecific assemblage. However, when looking at the skeletal completeness, we may find that the sample is more heterogenous in its preservation, with one Edmontosaurus having a completeness score of 40, and the rest of them having an average score of 4, and that the Troodon findings have a very low completeness of 0.2; this accumulation looks more like a monospecific assemblage with allochthonous elements transported from upstream. If we take into account the phylogenetic estimator and find out that most of the elements of Edmontosaurus have only a couple or none of the characters that test homologous features in Hadrosauromorpha, then it means that the remains may have been given the name by association to the most complete specimen with a completeness score of 40, and then it is best to treat this as potentially a multispecific assemblage.

The degree of taxonomic resolution in fluvial environments seems to be highly variable and not necessarily dependent on the energy level. While unranked material is as common as specimens identifiable to the family level, resolution fluctuates significantly when broken down into specific sub-environments. In this context, there are specimens identified by specific synapomorphies that allow for placement within the hierarchical taxonomy or phylogenetic arrangement, yet lack the diagnostic features required for a lower-level taxonomic assignment. Finally, this information can inform taxonomic decisions. For instance, if an assemblage is more likely to be allochthonous, this can impact species delineation by allowing morphologically similar species be interpreted as separate species coming from different environments or niches that were deposited in the same place due to the similar hydrodynamic properties of their bones.

5 Conclusion

The phylogenetic estimator pê and the skeletal completeness metric could be incorporated into datasets when reporting collection monographs or taxonomic lists to indicate users of the potential in a collection when performing large data analysis, making it easier to exclude or include datapoints, or even when planning collection visits in a budget. For instance, the entry Diplodocimorpha associated with specimen IGM 11514 and Titanosauria associated with IGM 6080 could be reported along with a set of parameters that enable easy comparison between them. Specimen IGM 11514 has a pê_D+OTS=21.099, whereas IGM 6080 has a pê_D+OTS=1.030 when compared to each other in the same dataset. The skeletal completeness metric can then be reported as a set of absent/present coordinates for (cranial, axial, appendicular) skeletal elements, so that IGM 11514 has a skeletal completeness coded as (0,0,1), whereas IGM 6080 has a skeletal completeness coded as (0,1,0), allowing a quick idea of the shape of the bones — the preserved diplodocoid remains are cylindrical long bones, whereas the titanosaur remains are represented by vertebrae; IGM 11514 was deposited in a prodelta environment, whereas IGM 6080 was recovered from an indeterminate fluvial environment, making the shape and type of bones relevant to have an idea of the hydrodynamic properties in the environment. The goal of the estimator is not to infer biological information, but to communicate parameters. In the dataset, the unnumbered specimen reported as Titanosauria MPD with a pê_D+OTS*=12.696 and skeletal completeness (0,0,1), can be quickly compared against IGM 6080 and help informing decisions about the dataset in a similar way achieved by other estimators, like average or standard deviation, before seeing the bioclasts.

Phylogenetic information moves from a specimen to a collection to a paper to a dataset, and in these consecutive filters, there is a loss of information due to the difficulty of coding it in a standardized and logical way. The estimator developed here does not pretend to provide any measurement on the object per se, but to communicate the research and collection bias surrounding each occurrence in order to set more objective exclusion and inclusion criteria. Reporting the skeletal completeness metric and phylogenetic informativeness in datasets have the potential to inform on taphonomic processes by highlighting the preferential preservation of specific bone shapes. Current scholarship prioritizes material that is complete or phylogenetically diagnostic, and fragmentary remains go unaccounted or superficially mentioned. Skeletal completeness and shape of the bioclasts is seldom coded in datasets, representing a missed opportunity for a more granular understanding of vertebrate fossil records. The phylogenetic estimator pê developed here aims to bridge the gap between the physical fossil record and the abstracted fossil record in the form of data points in a dataset by incorporating information that describes the apparent usefulness of a specimen in a standard, easy to compute way that can be readily retrieved and interpreted in a dataset.

Correction note

A correction has been made to this article. Details can be found at: 10.3389/fevo.2026.2014385.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. Standard features of Grammarly were used to assist with proofreading, but no Premium Features (https://www.grammarly.com/blog/product/how-grammarly-uses-ai/).

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fevo.2026.1882629/full#supplementary-material

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Footnotes

Edited by: Mauricio Omar Velázquez-Aguilar, Autonomous University of the State of Hidalgo, Mexico

Reviewed by: Victor Manuel Bravo-Cuevas, Autonomous University of the State of Hidalgo, Mexico

Arindam Chakraborty, Universiti Malaya, Malaysia

Simon Darroch, Vanderbilt University, United States

Atzcalli Ehécatl Hernández Cisneros, Universidad Autónoma de Baja California, Ensenada, Mexico