Osteohistological description of ostrich and emu long bones, with comments on markers of growth

Abstract

Ostriches and emus are among the largest extant birds and are frequently used as modern analogs for the growth dynamics of non‐avian theropod dinosaurs. These ratites quickly reach adult size in under 1 year, and as such do not typically exhibit annually deposited growth marks. Growth marks, commonly classified as annuli or lines of arrested growth (LAGs), represent reduced or halted osteogenesis, respectively, and their presence demonstrates varying degrees of developmental plasticity. Growth marks have not yet been reported from ostriches and emus, prompting authors to suggest that they have lost the plasticity required to deposit them. Here we observe the hind limb bone histology of three captive juvenile emus and one captive adult ostrich. Two of the three juvenile emus exhibit typical bone histology but the third emu, a 4.5‐month‐old juvenile, exhibits a regional arc of avascular tissue, which we interpret as a growth mark. As this mark is not present in the other two emus from the same cohort and it co‐occurs with a contralateral broken fibula, we suggest variable biomechanical load as a potential cause. The ostrich exhibits a complete ring of avascular, hypermineralized bone with sparse, flattened osteocyte lacunae. We identify this as an annulus and interpret it as slowing of growth. In the absence of other growth marks and lacking the animal's life history, the timing and cause of this ostrich's reduced growth are unclear. Even so, these findings demonstrate that both taxa retain the ancestral developmental plasticity required to temporarily slow growth. We also discuss the potential challenges of identifying growth marks using incomplete population data sets and partial cortical sampling.

Capacity for growth plasticity, ubiquitous in Vertebrata, is to be expected in ostriches and emus. Even so, because they typically reach skeletal maturity within one year and therefore lack annual markers of decreased osteogenesis, our study is the first to report that these large extant ratites still maintain the ability to reduce growth rates in response to external or physiological stimuli.

1. INTRODUCTION

Osteohistological studies of extant aves, and in particular the paleognaths, illuminate aspects of extinct dinosaur biology previously thought unattainable, like metabolic regime (Castanet et al., 2000; Cubo et al., 2016), biomechanics (de Ricqlès et al., 2000), and even fossil preservation rate (Marshall et al., 2021). These insights are predicated on studies demonstrating a one‐to‐one correspondence between arrested growth lines and chronological age (e.g., de Buffrenil & Castanet, 2000; Kohler et al., 2012; Montes et al., 2007). Even so, there is a growing chorus of studies reporting difficulties in differentiating between annually deposited, accurate skeletochronological indicators, and growth marks associated with exogenic factors (Heck & Woodward, 2021; Schucht et al., 2021; Starck & Chinsamy, 2002).

Paleognath birds are a group of primarily large and flightless taxa, which include ratites such as ostriches and emus. The exact phylogenetic position of these taxa within Palaeognathae is the subject of ongoing debate, but consistently ostriches are recovered as one of the earliest‐diverging extant palaeognaths, whereas emus are recovered as one of the latest‐diverging extant palaeognaths (Mitchell et al., 2014). Ratites are of interest to paleohistologists because, like their non‐avian dinosaur ancestors, these rapidly growing birds exhibit vascular architectures that span the full range between slowly deposited longitudinal vascularization and rapidly deposited radial vascularization. In extant ratites, these architectures can be correlated with long bone osteogenesis rates ranging from 20 to 80 μm/d (Castanet et al., 2000). Because non‐avian dinosaurs like Maiasaura and Tyrannosaurus exhibit comparable vascular architectures, it follows that these taxa and their relatives were able to sustain osteogenesis rates equivalent to avians (Cooper et al., 2008; Horner et al., 1999; Padian et al., 2001; Woodward et al., 2015). Unlike their extant avian descendants, most non‐avian dinosaurs required multiple years to attain adult size, indicated by the presence of annually deposited arrested growth rings within the bone cortex (de Buffrénil et al., 2021). In some dinosaur taxa such as Plateosaurus (Klein & Sander, 2007) and Tyrannosaurus (Woodward et al., 2020) variable spacing between annual cortical growth rings reflects plasticity in osteogenesis, possibly in response to nutrient availability (Cullen et al., 2014; Woodward et al., 2020). Despite their rapid osteogenesis, extant paleognaths remain smaller than many of their non‐avian dinosaur ancestors because their growth duration is significantly shorter; adult size is reached within 180–190 days (Cooper, 2005). As a byproduct of this truncated growth window, ostrich, and emu bones do not typically exhibit annually deposited arrested growth rings. Because of this, it is currently unclear if ostriches and emus have retained the ability to arrest growth and deposit growth marks, or if this ancestral plasticity is lost (Starck & Chinsamy, 2002). Here, we describe for the first time growth marks discovered in one ostrich and one emu and discuss the implications of their presence.

2. MATERIALS AND METHODS

2.1. Specimens used

For this study, we borrowed a set of transverse diaphyseal thin sections from a 3‐year‐old adult male ostrich, from the collections of the Museum of the Rockies (MOR 1707). The two samples each spanned two slides due to the large size of the femur in transverse section. In total, this produced four petrographic thin sections, two from each sample. One set was stained using Toluidine blue, the other was not. The ostrich was farm‐raised in Montana, but no additional information about its provenance or life history accompanied the specimen.

All three emu specimens hatched within the same year at the Montana Emu Ranch (Kalispell, MT). The individuals died on the ranch prematurely of unknown causes at 3.5, 4.5, and 5.5 months of age. The emu cadavers had been frozen and stored, and eventually hind limbs and sacra were shipped to OSU‐CHS overnight to avoid thawing. Specimens are accessioned into the Sam Noble Oklahoma Museum of Natural History osteology collections under specimen numbers OMNH RE 864, OMNH RE 865, and OMNH RE 866 respectively. Transverse mid‐diaphyseal blocks of right femora, tibiae, and fibulae were prepared as ground sections. The right fibula of the 4.5‐month‐old emu (OMNH RE 865) exhibits a fracture callus, but all other emu bones appear healthy. Because of this, slides were also prepared from the left side of the 4.5‐month‐old emu (OMNH RE 865) to control for potential variance induced by the pathology found in the specimen's right fibula. All specimens died prior to their donation for this study, so approval from the Institutional Animal Care and Use Committee was unnecessary (Oklahoma State University Policy 1–0505).

2.2. Methods

Because the specimen was already prepared by an unknown investigator for an unknown reason, the methods used to produce the ostrich petrographic thin sections (MOR 1707) are also unknown. Emu bones were prepared using methods established in Schweitzer et al. (2007) and modified in Woodward et al. (2014). In summation: specimens were skeletonized, preserved in a solution of 10% buffered formalin, dehydrated in a 70%, 85%, and 100% ethanol series over the course of 1 week, embedded in polyester resin, wafered and mounted, and ground into petrographic thin sections ranging between 20 and 100 microns in thickness. We also report that the ostrich slides range in thickness between 90 and 230 μm. All slides were imaged using a Nikon DS‐Ri 2 camera mounted to a Nikon Eclipse petrographic microscope with 2x and 5x objectives and an ASI automated stage. Slides were imaged under cross‐polarized, full‐wave plate, or plane‐polarized light. Photomosaic images were assembled using Nikon Elements: Documentation version 5.20.02, and figures were compiled in Adobe CC Photoshop and Illustrator. Once imaged, MOR 1707 ostrich sections spanning two slides were reassembled using Photoshop. Roughly 1 mm of material was lost when blocks were subdivided via tile saw to be mounted across multiple slides, so a 1 mm gap was added between photomosaic images when they were combined in Photoshop. Once fully reconstructed, the section was digitally traced in Photoshop, and the silhouette was imported into ImageJ, a product of Fiji (Schindelin et al., 2012). Using the BoneJ plugin (Doube et al., 2010), histomorphometric data like the centroid, circumference, and cross‐sectional area of cortical tissues were taken. Centroid coordinates were transferred to Photoshop, where measurements were taken.

2.3. Terminology

To facilitate the accurate identification of a growth mark, we use the following three qualitative diagnostic criteria: (1) parallel or lamellar fiber orientation that displays high anisotropy under cross‐polarized light, (2) reduced density and cross‐sectional area of osteocyte lacunae, and (3) radially flattened osteocyte lacunae and canaliculi. These criteria are qualitative and comparative, meaning that they are not explicitly measured, but rather described in comparison to the tissues deposited before and after it.

Once identified in a general sense, the “clarity” of growth marks can delineate them into two broad categories: Lines of Arrested Growth (LAGs) and annuli. A LAG presents as a distinct line, while an annulus presents as a diffuse ring of avascular tissue (Francillon‐Vieillot VdB et al., 1990). Note that “annulus” here refers to the ring‐like shape of the structure, as opposed to potential annual deposition of the structure. Although LAGs and annuli are presented herein as two distinct categories for descriptive purposes, these terms exist as the end points of a spectrum (Atterholt et al., 2021), and the clarity of growth marks can transition from one end of the spectrum to the other, depending on localized variance in growth rate. The standard physiological interpretation of these marks is that LAGs represent the complete cessation of growth, while annuli represent a period of slowed growth (Francillon‐Vieillot VdB et al., 1990). Ultimately, growth marks are delineated into these categories for descriptive precision, but the physiological significance of their identification (i.e., the demonstration of slowed growth and developmental plasticity) is the same for both categories.

To discuss the distribution of growth marks along the radius of bone shafts, the terms “cyclical growth marks” (CGMs) and “non‐cyclical growth marks” (NCGMs) are typically used, for which Padian and Lamm (2013) offer this definition: “CGMs can be distinguished from noncyclical marks because the appearance of the latter tends to be haphazard rather than regular (i.e., they do not reflect a particular spacing or rhythm) and the latter tend not to encircle the entire shaft but tend to be locally confined to an arc.” (p. 196). Interpretation of unevenly spaced growth marks as non‐cyclical may be incorrect as alligator osteohistological studies have demonstrated that uneven spacing may result from variable duration of growth hiatuses as opposed to irregular (i.e., non‐cyclical) timing (Woodward et al., 2014). Regardless, both growth marks described herein occur in isolation, so it is not possible to determine if they are evenly spaced relative to other marks. Regardless of their radial distribution, growth marks typically present as fully enclosed circles, but due to cortical remodeling and anisometric growth, they can also present as incomplete arcs.

Transverse, mid‐diaphyseal sections of adult animals often exhibit other markers of variable growth, like the outer circumferential layer (OCL), inner circumferential layer (ICL), and bright lines, all of which must be distinguished from growth marks to facilitate their accurate identification. During deposition of fibrolamellar cortical tissue, hypermineralized bright lines are formed, which can sometimes be erroneously misidentified as the hypermineralization associated with growth marks. Unlike growth marks, these bright lines are highly conformable with surrounding laminar vasculature, and their deposition via static osteogenesis yields a disorganized woven fiber organization that is distinct from the lamellar fiber organization of growth marks deposited via dynamic osteogenesis (Prondvai et al., 2014; Stein & Prondvai, 2014).

The OCL and ICL are both deposited as growth slows in senescence, but the diagnostic criteria used to identify and distinguish them differ slightly. Differentiating the ICL from cortical tissue (and by extension potential growth marks) is straightforward because the ICL deposits onto a resorptive surface that lines the medullary cavity, thus producing a non‐conformable “tide line” (Francillon‐Vieillot VdB et al., 1990, p. 505). In contrast, the OCL is always found beneath the periosteal surface, but differentiation of the OCL from appositional cortical tissue is more challenging because underlying resorptive surfaces are not always present. In the case of rapidly growing vertebrates such as large ratites and non‐avian dinosaurs, the OCL can generally be distinguished from appositional primary tissue by a pronounced reduction in relative vascular density in the outermost cortex. This region consists of slowly deposited parallel‐fibered or lamellar tissue with an increasingly tight spacing of growth marks (Cullen et al., 2021; Woodward et al., 2020) but in the absence of a resorptive surface, drawing a sharp “line” between cortical tissue and the OCL requires the alignment of multiple lines of evidence, as opposed to the identification of a single line of direct evidence.

For more general descriptions of these structures and others not explicitly described above, we follow standard terminology put forth by Francillion‐Viellot VdB et al. (1990, pgs 509–512), Chinsamy et al. (2013), and O'Connor et al. (2018).

3. RESULTS

3.1. 4.5‐month‐old emu

The juvenile emu tibia of OMNH RE 865 shows cortical thickness that varies from 2.6 mm along the anterior aspect of the cortex to 5.3 mm along the posterolateral aspect (Figure 1a). The endosteal surface is mostly smooth and uninterrupted, except for large nutrient foramina along its posterolateral border (Figure 1b). Tissue fibers appear isotropic under cross‐polarized light, suggestive of woven tissue (Figure 1c). In the anterior innermost cortex, a distinct sliver of compact coarse cancellous bone (CCCB) (Enlow & Yaeger, 1963) measuring approximately 1.9 mm thick can be seen (Figure 1d). These regions show increased reticular vascularization and enlarged osteocyte lacunae (70 μm vs 100 μm). Fiber bundles in this region are isotropic, reflecting their disorganization (Figure 1e). Vascularization remains reticular along the full thickness of the lateral portion of the cortex. Vascular orientation across the rest of the cortex is highly variable but ranges from reticular at the inner cortex to sublaminar toward the periosteal surface (Figure 1f). Wherever sublaminar primary vascularization is present across the section, locally hypermineralized ‘bright lines’ are present within tissue laminae (Figure 1g). Periosteal surface topography is continuous along most of its circumference, with sparse irregular vascular canals open to the surface (Figure 1h).

FIGURE 1.

Mid‐diaphyseal section of 4.5‐month‐old emu tibia, imaged under multiple light conditions, showing various structures discussed in “Results” section. (a) Photomosaic scan imaged under 20x magnification with anatomic directions. (b) Photomosaic of endosteal surface imaged under 20x magnification and cross‐polarized light, (c) CCCB/cortex interface imaged under 50x magnification and full‐wave plate light, (d) CCCB/cortex interface imaged under 50x magnification and full‐wave plate light, (e) CCCB imaged under 10x magnification and full‐wave plate light, (f) photomosaic of reticular to sublaminar vascularization imaged under 20x magnification and cross‐polarized light, (g) bright lines imaged under 100x magnification and cross‐polarized light, and (h) photomosaic of periosteal surface imaged under 20x magnification and polarized light

A growth mark is clearest along the anterior side of the cortical bone, between 260 μm and 350 μm from the periosteal surface (Figure 2a). It cannot be readily traced completely around the bone circumference. The terminal edges of the mark are ill‐defined, and show no scalloping indicative of active secondary resorption. The mark is clearest under plane light, where a modest darkening of the bone is present (Figure 2b). Under a full‐wave plate (Figure 2c) and cross‐polarized light (Figure 2d), changes in brightness and fiber orientation are inconsistently present along its length. The most striking feature of the mark is a change in osteocyte lacunar density and morphology. Osteocyte lacunae are oblong here, with their long axes aligned parallel to the periosteal surface. Communicating canaliculi between osteocyte lacunae are sparser in the regions surrounding the structure, and the channels that are present are more likely to connect adjacent osteocytes (Figure 2e). On aggregate, this localized osteocyte variance gives the mark its definition on a macroscopic scale. Radial offshoots of reticular vascularization terminate in contact with this horizon (Figure 2f). Sparse longitudinal and latitudinal vascular canals run parallel to the mark (Figure 2f).

FIGURE 2.

Mid‐diaphyseal section of 4.5‐month‐old emu tibia, imaged under multiple light conditions. (a) Photomosaic with anatomic directions. Black region is a preparation artifact caused by partial delamination of specimen from the slide. Red box indicates region enlarged in b–d. (b–d) photomosaic of the growth mark imaged under 20x magnification and imaged under (b) plane‐polarized, (c) full‐wave plate, and (d) cross‐polarized light. (e) Osteocyte lacunae at the growth mark, imaged under 100x magnification and polarized light and (f) osteocyte lacunae at the growth mark, imaged under 100x magnification and cross‐polarized light.

3.2. Three‐year‐old ostrich

The adult femur of MOR 1707 shows a cortical thickness between 3.9 mm and 5.8 mm, with its thickest region corresponding to trochanter minor (Vijayan et al., 2019) along the posteromedial aspect of the bone (Figure 3a). Another, more modest thickening is present along the anterolateral surface, and likely corresponds with the anterior intermuscular line (Figure 3a) (Vijayan et al., 2019). The endosteal surface is continuous, except for two prominent trabeculae located anteriorly and extending into the medullary cavity (Figure 3a). A complete, non‐conformable layer of avascular tissue separates the innermost cortex from the medullary cavity (the inner circumferential layer, ICL), and averages about 200 μm in thickness (Figure 3b). Sparse vascular canals radially intrude through the ICL, providing communication between the medullary cavity and the cortex, the region is otherwise avascular (Figure 3b). Under cross‐polarized light, the ICL exhibits anisotropic birefringent fiber orientation (Figure 3b).

FIGURE 3.

Mid‐diaphyseal section of femur from 3‐year‐old ostrich, imaged under multiple light conditions, showing various structures discussed in “Results” section. (a) Photomosaic scan imaged under 20x magnification and plane‐polarized light, with anatomic directions. (b) ICL/cortex interface, imaged under 20x magnification and cross‐polarized light. (c) Photomosaic scan of toluidine‐blue stained slide, imaged under 20x magnification and plane‐polarized light, note sublaminar to laminar vascularization throughout the cortex. (d) Photomosaic was taken at 20x magnification under cross‐polarized light to illuminate hypermineralized vascular laminae. (e) Sublaminar vascularization with sparse woven matrix, imaged under 50x magnification and waveplate retardation. (f) Photomosaic scan of toluidine‐blue stained slide, imaged under 20x magnification and polarized light. Note reticular vascularization deep to trochanter minor. (g) Outer circumferential layer (OCL) imaged under 50x magnification and cross‐polarized light

Throughout most of the cortex, vascular canal orientation is laminar, though areas of sublaminar tissue are sparsely present along the inner regions of the cortex (Figure 3c). Secondary osteons are absent throughout the cortex. Under cross‐polarized light, regions with laminar and sublaminar vascularization also exhibit localized lines of hypermineralized matrix that lie at the midpoint between primary laminar vascularization (Figure 3d). Unlike hypermineralization typically seen in growth marks, these lines undulate alongside local primary vascularization (Figure 3d). These lines also exhibit disordered fiber direction, suggesting that they are composed of woven tissue (Figure 3e).

Wide, reticular vascularization comprises trochanter minor throughout the full thickness of the cortex (Figure 3f). These regions are also non‐conformable with surrounding vascularization, suggesting that the area experienced secondary remodeling during ontogeny (Figure 3f). Likewise, the cortex consisting of the anterior intermuscular line shows vascularization that is predominantly sublaminar with isolated gradations to subreticular vascularization. Volkmann's canals in this region become wider in diameter and radially elongated adjacent to the feature's surface (Figure 3f). A thick layer of avascular, lamellar tissue approximately 150 μm thick completely encloses the outer cortex (outer circumferential layer, OCL) (Figure 3g).

A growth mark can be traced around the entire circumference of the bone (Figure 4a). It travels largely uninterrupted within heavily remodeled regions like the reticular vascularization forming trochanter minor (Figure 3f) and is only occasionally interrupted by sparse Volkmann's canals (Figure 4a). The morphology of the mark is ovate, ranging from a 1.9 cm radius along its minor axis to a 2.6 cm radius along its major axis. Cortex deposition from the mark to the periosteal surface is consistently 1.4 mm in thickness (Figure 4a). Under plane‐polarized light, the mark is slightly lighter in appearance (Figure 4b). Stained with Toluidine blue, adjacent tissue is slightly lighter in plane light. Under cross‐polarized (Figure 4c) and waveplate‐retarded light (Figure 4d), the mark is modestly brighter and fiber orientation is consistently distinct from surrounding tissue. Osteocyte lacunae within the structure are sparse, radially flattened, and less interconnected by communicating canalicular channels (Figure 4b).

FIGURE 4.

Mid‐diaphyseal section of MOR 1707, a femur from 3‐year‐old ostrich. Photomosaic shows the growth mark imaged under multiple light conditions and magnifications. (a) Photomosaic scan of entire specimen, with region in red box expanded to see the potential mark. Imaged under 20x magnification and plane‐polarized light. Red box in the expanded region indicates regions imaged by b‐d. red arrows indicate mark. (b‐d) growth mark imaged at 100x magnification and various light conditions. Red arrows indicate the mark. (b) Mark imaged under polarized light. Note flattened osteocyte lacunae and canaliculi. (c) Mark imaged under cross‐polarized light. Note inconsistent hypermineralization of the region. (d) Mark imaged under cross‐polarized light and wave‐plate retardation. Note distinct fiber orientation

4. DISCUSSION

4.1. Emu interpretations

With the exception of the mark found in the 4.5‐month‐old emu tibia, all other bone histology is typical of rapidly growing juvenile emus (Cooper, 2005). In this tibia section, the mark shows modest hypermineralization, parallel‐fibered bone, sparse vascularization, and flattened osteocyte lacunae. As expected, there is no ICL or OCL in this juvenile, so this avascular tissue cannot be associated with these layers. Likewise, this structure does not undulate with nearby primary laminar vascularization, and so cannot be described as a bright line. Because this mark meets our diagnostic criteria and cannot be identified as any alternatives, we identify it as a growth mark. This mark presents as a distinct line as opposed to a diffuse ring and therefore fits the anatomical diagnostic criteria of a LAG. LAGs are typically interpreted to represent full cessation of growth, and while we agree with this interpretation, we note that this cessation was highly localized to a regional arc. The edges of the mark are diffuse until no longer distinguishable but exhibit no obvious interruptions due to resorption or primary vascularization. There are also no other marks present within the cortex, so we cannot comment on its potential as a cyclically deposited skeletochronological indicator. Because the animal is only 4.5 months of age, this mark was unlikely to be deposited as part of an annual fluctuation in osteogenesis rates. Furthermore, this mark is not present in the femur or contralateral tibia, which does not support physiologically induced systemic growth cessation seen in, for example, alligators (Woodward et al., 2014).

Vertebrates undergoing anisometric cortical growth often produce cortical drift lines similar to this (Enlow & Yaeger, 1963), but the absence of this mark in the other emus suggests that it formed not as a byproduct of standard ratite growth dynamics, but rather as a product of individual pathological history. Unlike the two other individuals raised alongside it, the 4.5‐month‐old emu presented with a broken left fibula and a growth mark in its right tibia, but not in corresponding contralateral elements. Biomechanically adaptive bone modeling has been hypothesized as a response to broken right fibulae in two juvenile Maiasaura (Cubo et al., 2016) but these presented as distinct crescents of radial fibrolamellar tissue in the adjacent right tibia. Alternatively, biomechanical stress associated with fibular injury may have caused a temporary arrest of growth in the emu. The co‐occurrence of this growth mark with pathology provides only circumstantial evidence, and a controlled experiment is required for testing this hypothesis.

Regardless of our growth mark interpretations, its inconsistent presence underscores the importance of using complete transverse sections and controlling for individual outliers by examining multiple specimens. If this thin section was prepared from only a core drilled from the cortex, as is sometimes done in paleohistological studies, then this mark could be interpreted as a complete ring, as opposed to an incomplete arc. Likewise, if presumably uninjured emu individuals were not included in this dataset, then the pathological presentation of a growth mark could be erroneously interpreted as a population‐wide phenomenon.

4.2. Ostrich interpretations

With the exception of the growth mark, the osteohistology described herein is as expected for an adult ostrich, showing predominately laminar to sublaminar vascularization, woven fiber matrix, an OCL, and an ICL. Because the mark shows (1) hypermineralization, (2) parallel fibered bone, (3) sparse vascularization, and (4) flattened osteocyte lacunae, we confidently identify it as a growth mark. A clear resorptive surface distinguishes the ICL from primary cortical tissue (Figure 3b), so we do not identify the mark as being part of the ICL. A non‐conformable resorptive surface that separates the OCL from the cortex is not consistently present, but a thick layer of fibrolamellar tissue separates the mark from incontrovertible OCL tissue (Figure 3c,d), and so we feel confident that the mark is not part of an early onset of the OCL. The mark is also non‐conformable with surrounding vasculature and does not show a woven fiber organization, which distinguishes it from hyperminerized bright lines. The growth mark is occasionally interrupted by primary vascularization, but it is still identifiable as a complete ring. Some regions of the mark closely resemble a clearly defined LAG, but most regions present as a diffuse annulus. From this, we interpret that growth slowed in most regions along the cortex, and entirely stopped in others. We hypothesize that due to its load‐bearing role, the presence of a growth mark in the femur is reflective of system‐wide slowed growth, but cannot test this hypothesis, as additional skeletal elements for the individual are unavailable. Furthermore, because we observe only a single mark, we cannot comment on the potential periodicity of growth arrest. Additionally, Schutch et al. (2021) have demonstrated that the final number of visible growth marks may be at least partially determined by methods with which the specimen was prepared, and so a slide prepared via microtome sectioning may yield more marks that are currently invisible in our data. Unfortunately, because we lack the original, undehydrated specimen, this study cannot be undertaken.

The unknown individual life histories of these specimens preclude the definitive identification of underlying causes, but the presence of a growth mark within the cortex of an ostrich and emu, reported here for the first time, remains significant. Our findings add ostriches and emus to the list of Aves that retain an ancestral non‐avian dinosaur capacity to temporarily slow or stop osteogenesis, which also includes Diatryma, Amazon amazonica, the New Zealand moa, and the New Zealand kiwi bird (Bourdon et al., 2009; Heck & Woodward, 2021). Due to the infrequency with which growth marks are observed and reported in Neornithine taxa, further investigation is required to pinpoint their cause(s).

5. CONCLUSIONS

Growth plasticity can be observed in all vertebrates (de Buffrénil et al., 2021). More specifically, growth marks have been reported in crocodilians (Woodard et al., 2014), non‐avian dinosaurs, extant neognaths (De Ricqlès et al., 2001), paleognathous kiwi birds (Heck & Woodward, 2021), moa (Turvey et al., 2005), and the elephant bird (De Ricqlès et al., 2016). Thus, capacity for growth plasticity, ubiquitous in Vertebrata, is to be expected in ostriches and emus. Even so, because they typically reach skeletal maturity within 1 year and therefore lack annual markers of decreased osteogenesis, our study is the first to report that these large extant ratites still maintain the ability to reduce growth rates in response to external or physiological stimuli.

AUTHOR CONTRIBUTIONS

H. N. W. conceived the experiment, prepared emu slides, and reviewed figures, tables, and drafts of the paper, and supplied the reagents, materials, and analysis equipment. N. S. O. formatted the manuscript, performed the experiment for the emu individuals, analyzed their data, wrote the paper, prepared figures and/or tables, and reviewed drafts of the paper. B. H‐F. performed the experiment for the ostrich individual and analyzed the data, wrote the paper, aided in figure preparation, and reviewed drafts of the paper. K. T. edited a manuscript draft, provided initial identification of ostrich LAG, and provided initial digital imaging for analyses.

Ong, N. , Hart‐Farrar, B. , Tremaine, K. & Woodward, H.N. (2022) Osteohistological description of ostrich and emu long bones, with comments on markers of growth. Journal of Anatomy, 241, 518–526. Available from: 10.1111/joa.13665

DATA AVAILABILITY STATEMENT

N/A