The life histories of Palaeocene mammals are poorly known, but may have been central to their success in diversifying across terrestrial ecosystems after the end-Cretaceous extinction. Among these mammalian groups, the eutherian Taeniodonta are particularly enigmatic, with few modern analogues and no living descendants, despite being one of the only lineages to apparently traverse the Cretaceous-Palaeogene (K-Pg) boundary. Here we investigate the life history of an early Palaeocene taeniodont, Conoryctes comma, based on a multi-individual, multi-element sample. We produced palaeohistological thin sections of skeletal elements, including one tooth and 26 postcranial bones, to produce 36 slides. Nearly all elements sampled exhibit similar osteohistological architecture, with a small internal zone of compacted coarse cancellous bone surrounded by an internal cortex of periosteally-derived fibrolamellar bone of variable thickness, and an outer cortex of lamellar bone. The well-vascularized fibrolamellar complex in the limb bones, lacking cyclical growth marks, is indicative of overall rapid growth to near adult body size. Cyclical growth marks are present in the outer cortex after the transition to slow-growing lamellar bone, but not in the inner cortex, suggesting sexual maturity was reached in one year. In some elements, an internal growth mark shares histological similarities with weaning marks in living mammals and other contemporary Palaeocene mammals, and occurred at the body size predicted for this transition in therian mammals. The unusual presence of compacted coarse cancellous bone near the midshafts of multiple limb bones may be related to cortical thickening, and is similar to the arrangement described in some fossorial mammals, supporting previous assertions of this lifestyle in Conoryctes. Altogether, these palaeohistological signals suggest a life history in Conoryctes comma similar to living eutherians, despite uncertainty about whether it is within crown Placentalia or a close outgroup. Thus, our data are consistent with an early origin of placental-like reproductive strategies in their eutherian ancestors, although this attribute was likely shared more broadly among Mesozoic lineages prior to the end-Cretaceous extinction. These data may be useful in the future for broader evaluations of early Palaeocene mammal growth and life history, or for other physiological proxies.

Specimens:

We selected a range of postcranial skeletons of Conoryctes comma, all collected from the Nacimientio Formation in the San Juan Basin of New Mexico (Fig. 1). In this paper, we consider Conoryctes to likely be synonymous with, or very closely related to, another named taeniodont genus, Huerfanodon, following Kynigopoulou (2023). Both taxa overlap in time and are nearly identical in size, being distinguished only by relatively minor differences in premolar morphology (e.g., presence/absence of P4 protocone; presence/absence of p4 metaconid), which is highly variable even between members of the same species (Kynigopoulou, 2023). However, it is outside the scope of this study to formally synonymize these names, and this endeavour is in progress elsewhere. Where the postcranial material in our sample is associated with dental material (e.g., NMMNH P-48198), the teeth are referrable to Conoryctes comma (Kynigopoulou et al., 2024). Further, because duplicate postcranial elements are identical in morphology and palaeohistological growth signals (i.e., they do not exhibit different growth trajectories), we consider them all to represent a single taxon, Conoryctes comma. The San Juan Basin where the specimens were collected is the historic territory of the Ute and Navajo (Diné) peoples, and all specimens were collected under permit from public lands administered by the United States Bureau of Land Management. All specimens are accessioned at the New Mexico Museum of Natural History and Science (NMMNH). Specimens range in age from middle (To2) to late (To3) Torrejonian, spanning from about 62.7 to 62.2 Ma (Flynn et al., 2020; Leslie et al., 2018). Specimen details and parts of the specimens sampled are outlined in Table 1.      

            The sample aimed to include not only multiple individuals of varying sizes, for evaluating ontogenetic variation in tissue types, but also multiple elements from each individual, to assess intraskeletal histovariability. This is because bones from individuals of different ages record different windows of life, and bones from different regions of the skeleton can vary in how their local growth patterns reflect overall mass increase in the body. Thus, broad sampling is necessary when studying lineages with poor histological characterization, particularly because the elements whose growth records most closely reflect overall body mass increase are not necessarily consistent between clades (Cullen et al., 2014; Horner et al., 2000; Klevezal, 1996; Kolb et al., 2015; Montoya‐Sanhueza et al., 2021; Padian et al., 2016, 2013; Woodward, 2019; Woodward et al., 2015, 2014). In Conoryctes, like in other Palaeocene eutherians studied (Funston et al., 2022), we found minimal variation in the overall growth signal between bones (see discussion). Where possible, sampling targeted the minimum shaft circumference of each bone, because this is typically closest to the neutral growth zone and more closely reflects the overall growth of the individual (Enlow, 1963; Padian et al., 2013), particularly in simply cylindrical elements like the tibia. However, in many cases, this area was not preserved, the bones had more complex shapes, or the specimens available for sampling were small but identifiable fragments of the element. Accordingly, some sections had to be taken closer to the metaphyses, which typically have more complex growth signals that reflect not only size increase, but also shape change and soft tissue attachments. Likewise, for some sections only an approximate position of the sample relative to the entire element is known.

Specimens were photographed using a Nikon D850 camera with a 60 mm MicroNikkor lens prior to thin-sectioning. Some elements were digitized using photogrammetry in Agisoft Metashape Standard (version 1.6.1), to produce three-dimensional models for conservation of morphology.

            Specimens were thin sectioned following the protocol of Funston et al. (2022), which is a modified version of the method described by Lamm (2013). In brief, sample regions were removed from specimens either using natural breaks, or by stabilizing with paraloid and cutting on a Buehler Isomet 1000 Precision saw. A major difference from other thin-sectioning protocols is that sample regions were adhered in predetermined orientations to pre-cured epoxy bases using Loctite cyanoacrylate, to facilitate sampling in the correct plane. Specimens on bases were then embedded in Buehler Epothin II epoxy resin under a vacuum and left to cure for 24 hours. Embedded blocks were sectioned at the plane of interest using the Isomet 1000 saw, but unlike the protocol of Lamm (2013), specimens were kept as blocks, rather than serially cut into wafers, as this minimized the potential for loss of the friable bones and teeth. Cut block faces were consolidated using cyanoacrylate as necessary, before being frosted using 600 grit silicon carbide abrasive slurry. Unlike most thin-sectioning protocols (Cerda et al., 2020; Chinsamy and Raath, 1992; Cuccu et al., 2025; Lamm, 2013; Montoya-Sanhueza and Chinsamy, 2017), but critical for optimal adhesion of fragile dental tissues, frosted acrylic slides were inverted to mount onto the blocks using cyanoacrylate without applying any pressure. Typically, glass slides and epoxy adhesive are used, and specimens are weighed down or clamped for mounting, but this often temporarily warps the slide, which results in flexion and breakage of dental tissues when slides are polished to low thicknesses. Cyanoacrylate and acrylic make a stronger bond than glass and epoxy, because the acrylic ‘melts’ in contact with cyanoacrylate, merging into a single material, and thus the bond does not rely on the surface area adhesion of two materials (epoxy and glass). Acrylic provides a further advantage in that it is more resistant to shattering, prolonging the lifespan of the completed section. Mounted samples were resectioned using the Isomet 1000 saw to a thickness of 700 µm, and were then manually ground on a glass plate using 600- and 1200-grit silicon carbide abrasive slurries. Slides were ground until they reached the desired optical clarity, although for many specimens in the sample, taphonomic or diagenetic alteration of the material resulted in replacement of original bone matrices with extensive regions of opaque minerals. Because of the modified mounting protocol, these slides could be polished to very low thicknesses, in some cases < 15–30 µm thick, to maximize light transmission. Thus, some information was still accessible from these specimens using cross-polarized light, which would not have been available using standard palaeohistological methods.

            Slides were imaged using a Leica DMLP Transmitted Light Polarizing Microscope with Leica Application Suite 4. Photomontages were created using the automated ‘photomerge’ feature of Adobe Photoshop 2022. Images were taken under plane polarized light, cross-polarized light, and cross-polarized light with a 430 nm ‘lambda’ filter, which aids in distinguishing opaque regions from extinct regions under cross-polarized light.