Palaeontologia Electronica http://palaeo-electronica.org COMPARATIVE VARIABILITY OF INTERMEMBRANOUS AND ENDOCHONDRAL BONES IN PLEISTOCENE MAMMALS Kristina R. Raymond and Donald R. Prothero ABSTRACT Study of the embryology and ossification of modern bones predicts that fossil intermembranous bones (which ossify from connective tissue) will exhibit greater size variability than endochondral bones (which are formed from embryological cartilagi- nous precursors), because intermembranous bones are less tightly constrained by joints and articular surfaces. To evaluate this hypothesis, we measured multiple dimen- sions of 989 intermembranous bones (patellae and other sesamoids) of the saber- toothed cat Smilodon fatalis, the Ice Age lion Panthera atrox, the bison Bison antiquus, the horse Equus occidentalis, the camel Camelops hesternus, the ground sloths Paramylodon (=Glossotherium) harlani and Nothrotheriops shastensis from Rancho La Brea and from the late Pleistocene San Josecito Cave in Nuevo Leon, Mexico. These were compared to measurements of 811 endochondral bones (primarily astragali) of comparable size. Through statistical analyses (coefficients of variation, ANOVA, modi- fied Levene’s test, and t-tests) we found slight evidence of higher variability in many of the intermembranous bones of these taxa (21 out of 27 CVs were higher for intermem- branous bones than endochondral bones), although this trend is not found in all taxa. Using a modified Levene’s test, only Smilodon and some of the dimensions of horse and bison patellae are significantly more variable than the corresponding dimensions of the astragali. Although the results are mixed, at least some data show that inter- membranous bones are not as tightly constrained by growth and by adjacent tissues as are endochondral bones. This evidence of relative variability is important in assess- ing how much variability is typical of a single species, and thus has taxonomic implica- tions. Kristina R. Raymond, Don Sundquist Center for Excellence in Paleontology, Department of Biology, East Tennessee State University, Johnson City, TN 37614 USA. [email protected] Donald R. Prothero, Department of Geology, Occidental College, Los Angeles, CA 90041 USA. [email protected] KEY WORDS: intermembranous bone; endochondral bone; sesamoid bone; variability; Pleistocene PE Article Number: 13.1.4A Copyright: Society for Vertebrate Paleontology March 2010 Submission: 19 December 2008. Acceptance: 25 January 2010 Raymond, Kristina R. and Prothero, Donald R., 2010. Comparative Variability of Intermembranous and Endochondral Bones in Pleistocene Mammals. Palaeontologia Electronica Vol. 13, Issue 1; 4A: 14p; http://palaeo-electronica.org/2010_1/184/index.html RAYMOND & PROTHERO: VARIABILITY IN MAMMAL BONES INTRODUCTION Vickaryous and Olson (2007) point out that although sesamoids receive little attention in pale- In recent years, developmental biology has ontological literature, the majority of tetrapod lin- made enormous strides in understanding the eages develop at least one sesamoid. “As a group, growth and modification of bones, and the con- sesamoids and their ilk represent something of an straints on bone growth as well (Hall 2005; Currey anatomical enigma, with an enormous degree of 2006). Most paleontologists no longer view bones variability in size, shape, and position both within as static entities, but as dynamic three-dimensional and between taxa. Consequently, most skeletal objects that can vary in shape not only during descriptions relegate these elements to passages ontogeny, but also due to changed biomechanical that summarize… bones and cartilages, predispos- forces. Because of both ontogenetic and ecophe- ing them to continued marginalization” (Vickaryous notypic factors, bone shapes and sizes can vary and Olson 2007). This scientific neglect is largely quite widely in a population (Yablokov 1974). because sesamoids are not as commonly pre- Shape and size of fossil bones are important fac- served as more massive and larger bones of skele- tors in making taxonomic decisions, and assessing ton, or sometimes cannot be reliably associated the variability of a single population is critical in with a known species. deciding how much variability in a fossil sample Recent anthropological literature has com- can be attributed to a single species, or requires mented on patellar variability in humans of both other explanations. recent and Pleistocene age. Trinkaus (2000) con- In particular, variability due to different styles cludes that all of his samples “exhibit considerable of growth between endochondral bones (which variability in these patellar proportions.” Trinkaus ossify directly from an embryonic cartilaginous pre- and Rhoads (1999) and Ward et al. (1995) also dis- cursor, often constrained by joints and articular sur- cussed the variability of fossil hominid patellae, but faces) and less constrained intermembranous in the context of functional morphological interpre- bones, is highly relevant to these issues. The topic tations, rather than comparative variability. The lit- of intermembranous and endochondral bone erature cited above suggests that because modern growth, size and variability is one that is not com- lineages show patellar variability, the Pleistocene monly touched upon, except briefly in passing, in fossil record may also provide data to suggest that paleontological literature. Generally, intermembra- sesamoids have been variable throughout the his- nous bones are measured and discussed as only a tory of life. slightly relevant topic in regards to larger studies of Walmsley (1940), Bland and Ashhurst (1997), species or interspecific variation and sexual size and Bongers et al. (2005) discuss the development dimorphism. of the patella and its ossification from connective Intermembranous bones form directly from tissues and hyaline cartilage. Sarin et al. (1999) the connective tissue late in embryological devel- and Goldberg and Nathan (2004) analyzed the opment and after birth through intramembranous variability of human sesamoid bones, but without ossification. Some intermembranous bones, such comparing this variability to that of endochondral as the kneecap (patella), are almost always ossi- bones or using it in a systematic context. Our own fied in adult mammals (with minor exceptions). impetus for this research was stimulated when we Other intermembranous bones, known as sesam- noticed a similar high level of variability in ground oids, occur only in areas where a tendon passes sloth patellae (Figure 1), and the second author over a joint, and ossify in irregular and unpredict- has seen many similar instances in the large col- able patterns (Vickaryous and Olson 2007). The lections of fossils he has examined over the past number and shape of intermembranous bones vary 40 years. Prothero (2005, tables 5.1-5.9) docu- greatly within the Mammalia, and are highly taxon- mented some of this variability in North American dependent. Humans have only one sesamoid (the rhinocerotid patellae. pisiform) in the carpus. In many mammals, such Based on these considerations, intermembra- bones include the patella and large sesamoids in nous bones are predicted to show a higher level of the manus and pes. In ungulates, on the other variability than endochondral bones because they hand, the only large sesamoid element is the have limited articulation with other bones and are patella. The sesamoids in the manus or pes are formed through intramembranous ossification. small nodular ossifications in the digital flexor ten- Many intermembranous bones are referred to as dons, both at the metapodial-phalangeal joint and ‘free-floating.” By contrast, endochondral bones the distal interphalangeal joint; suids have as many are more constrained from unusual growth by artic- as 13 sesamoids in the manus alone. 2 PALAEO-ELECTRONICA.ORG FIGURE 1. Patellae (kneecaps) of the ground sloth, Paramylodon harlani, show apparent shape and size variability. Scale bar equals 1 cm. ulations with other bones. Apparent variability may Brea tar pits (Stock and Harris 1992) and cave also be the result of ossification into the tendons deposits such as San Josecito Cave in Nuevo because animals tend to replace minor tissue dam- Leon, Mexico (Arroyo-Cabrales et al. 2003) age with bone at the intersection of bone and ten- In this study, we measured samples of inter- don as an inflammatory response. Therefore, older membranous and endochondral bones from nearly individuals or individuals who have suffered tendon all the large mammals for which a sufficiently large or joint injury would display larger or oddly shaped sample (wherever possible, more than 100 of each sesamoids (Andrew Clifford, personal communica- element) exists. These include two ground sloths tion, 2007). (Paramylodon harlani and Nothrotheriops shasten- sis), the extinct bison Bison antiquus, the extinct MATERIALS AND METHODS horse Equus occidentalis, the lamine camelid Camelops hesternus, the Ice Age lion Panthera One of the best ways to address this question atrox, and the saber-toothed cat Smilodon fatalis. is to measure a relatively large set of fossilized The only common mammal not included in the intermembranous and endochondral bones from a study was the dire wolf (Canis dirus), which is cur- homogeneous population. Unfortunately, most fos- rently under study by F.R. O’Keefe (personal com- sil samples only rarely preserve sesamoid bones, mun., 2008). especially the small sesamoids of the manus and Sloths not only possess patellae