Archaeological and Anthropological Sciences

, Volume 11, Issue 9, pp 4663–4680 | Cite as

Classifying agency in bone breakage: an experimental analysis of fracture planes to differentiate between hominin and carnivore dynamic and static loading using machine learning (ML) algorithms

  • Abel MoclánEmail author
  • Manuel Domínguez-Rodrigo
  • José Yravedra
Original Paper


The analysis of bone breakage has always been underrepresented in taphonomic studies. Analysts, thus, lose the opportunity to resolve an important part of the equifinality related to activities that hominins and different types of carnivores may produce. Recent studies have shown that the use of powerful machine learning (ML) algorithms allow the accurate classification of bone surface modifications (BSM). Here, we present an experimental study, applying these algorithms to the analysis of bone breakage patterns. This statistical methodology allows the correct classification of three different assemblages which have been generated anthropogenically and by the activity of carnivores (i.e., hyenas and wolves). ML algorithms applied to a multivariate set of properties of broken bone specimens yielded an accuracy of 95% and were higher in classifying agency without the need to include information from BSM. This paper proposes a methodological approach that opens the door to improve our understanding of referential frameworks regarding bone breakage and to determine agency in prehistoric bone breakage processes.


Taphonomy Machine learning Algorithm Bone breakage Fracture planes 



AM is deeply grateful to the volunteers for their effort in helping with the anthropogenic bone breakage, particularly Alicia Caboblanco for her assistance during all the experimental work. We cannot forget the help provided by Cárnicas DIBE S.L. who provided all the carcasses that were used in the preparation of the anthropic broken bone assemblage. We also acknowledge Lloyd Courtenay for his initial assistance with the manuscript and Andrea Díaz Cortés for her last-minute comments. We thank Charles Egeland for his constructive comments in an earlier draft of this manuscript. The KND2 remains were collected during the course of unrelated archeological research in the Lake Eyasi basin, which was granted permission by the Tanzanian Commission for Science and Technology (COSTECH) and the Department of Antiquities and was funded by a National Science Foundation Doctoral Dissertation Improvement Grant (NSF-0620262) and a Wenner-Gren Foundation Dissertation Fieldwork Grant, both granted to Mary E. Prendergast. We thank her for allowing us to have access to the KND2 collection. We acknowledge Santiago Domínguez-Solera for his assistance in Hosquillo (Cuenca), as well as to the Parque del Hosquillo for his hospitality in allowing us to collect the samples studied in the research.

Funding information

AM is funded by a grant from the Junta de Castilla y León financed in turn by the European Social Funds through the Consejería de Educación (BDNS 376062).

Supplementary material

12520_2019_815_MOESM1_ESM.txt (127 kb)
ESM 1 (TXT 127 kb)


  1. Alcántara García V, Barba Egido R, Barral del Pino JM et al (2006) Determinación de procesos de fractura sobre huesos frescos: un sistema de análisis de los ángulos de los planos de fracturación como discriminador de agentes bióticos. Trab Prehist 63:37–45CrossRefGoogle Scholar
  2. Arriaza MC, Domínguez-Rodrigo M (2016) When felids and hominins ruled at Olduvai Gorge: a machine learning analysis of the skeletal profiles of the non-anthropogenic bed I sites. Quat Sci Rev 139:43–52. CrossRefGoogle Scholar
  3. Behrensmeyer AK (1978) Taphonomic and ecologic information from bone weathering. Paelobiology 4:150–162CrossRefGoogle Scholar
  4. Binford LR (1981) Bones: ancient men and modern myths. AcademicGoogle Scholar
  5. Blasco R, Domínguez-Rodrigo M, Arilla M, Camarós E, Rosell J (2014) Breaking bones to obtain marrow: a comparative study between percussion by batting bone on an anvil and hammerstone percussion. Archaeometry 56:1085–1104. CrossRefGoogle Scholar
  6. Blumenschine RJ (1988) An experimental model of the timing of hominid and carnivore influence on archaeological bone assemblages. J Archaeol Sci 15:483–502. CrossRefGoogle Scholar
  7. Blumenschine RJ, Selvaggio MM (1991) On the marks of marrow bone processing by hammerstones and hyaenas: their anatomical patterning and archaeological implications. In: Clarke JD (ed) Cultural beginnings: approaches to understanding early hominid life ways in the African Savanna. R. Habelt, Bonn, pp 17–32Google Scholar
  8. Booth JG, Sarkar S (1998) Monte Carlo approximation of bootstrap variances. Am Stat 52:354–357. Google Scholar
  9. Brain CK (1981) The hunters or the hunted?: an introduction to African cave taphonomy. University of Chicago Press, ChicagoGoogle Scholar
  10. Bunge M (1981) Analogy between systems. Int J Gen Syst 7:221–223. CrossRefGoogle Scholar
  11. Bunn HT (1983) Comparative analysis of modern bone assemblages from a san hunter-gatherer camp in the Kalahari Desert, Botswana, and from a spotted hyena den near Nairobi, Kenia. In: Clutton-Brock J, Crigson C (eds) Animals and archaeology: hunters and their prey. Oxford, pp 143–148Google Scholar
  12. Cáceres I, Bravo P, Esteban M et al (2002) Fresh and heated bone breakage. An experimental approach. In: De Renzi M, Pardo Alonso M, Belinchón M et al (eds) Current topics on taphonomy and fossilization. Ayuntamiento de Valencia, Valencia, pp 471–479Google Scholar
  13. Campmas É, Beauval C (2008) Consommation osseuse des carnivores: résultats de l’étude de l’exploitation de carcasses de bœufs (Bos taurus) par des loups captifs. Ann Paléontol 94:167–186. CrossRefGoogle Scholar
  14. Capaldo SD, Blumenschine RJ (1994) A quantitative diagnosis of notches made by hammerstone percussion and carnivore gnawing in bovid long bones. Am Antiq 59:724–748CrossRefGoogle Scholar
  15. Chernick MR (2007) Bootstrap methods: a guide for practitioners and researchers, 2nd edn. Wiley-Interscience, HobokenCrossRefGoogle Scholar
  16. Coil R, Tappen M, Yezzi-Woodley K (2017) New analytical methods for comparing bone fracture angles: a controlled study of hammerstone and hyena (Crocuta crocuta) long bone breakage. Archaeometry. 59:900–917. CrossRefGoogle Scholar
  17. de Juana S, Domínguez-Rodrigo M (2011) Testing analogical taphonomic signatures in bone breaking: a comparison between hammerstone-broken equid and bovid bones. Archaeometry 53:996–1011. CrossRefGoogle Scholar
  18. Domínguez-Rodrigo M (2015) Taphonomy in early African archaeological sites: questioning some bone surface modification models for inferring fossil hominin and carnivore feeding interactions. J Afr Earth Sci 108:42–46. CrossRefGoogle Scholar
  19. Domínguez-Rodrigo M (2002) Hunting and scavenging by early humans: the state of the debate. J World Prehist 16:1–54. CrossRefGoogle Scholar
  20. Domínguez-Rodrigo M (2018) Successful classification of experimental bone surface modifications (BSM) through machine learning algorithms: a solution to the controversial use of BSM in paleoanthropology? Archaeol Anthropol Sci doi:
  21. Domínguez-Rodrigo M, Baquedano E (2018) Distinguishing butchery cut marks from crocodile bite marks through machine learning methods. Sci Rep 8:5786. CrossRefGoogle Scholar
  22. Domínguez-Rodrigo M, Egido RB, Egeland CP (2007) Deconstructing Olduvai: a taphonomic study of the bed I sites. Springer Netherlands, DordrechtCrossRefGoogle Scholar
  23. Domínguez-Rodrigo M, Gidna A, Yravedra J, Musiba CM (2012) A comparative neo-taphonomic study of felids, hyaenids and canids: an analogical framework based on long bone modification patterns. J Taphon 10:147–164Google Scholar
  24. Domínguez-Rodrigo M, Martínez-Navarro B (2012) Taphonomic analysis of the Early Pleistocene (2.4 Ma) faunal assemblage from A.L. 894 (Hadar, Ethiopia). J Hum Evol 62:315–327CrossRefGoogle Scholar
  25. Domínguez-Rodrigo M, Pickering TR (2010) A multivariate approach for discriminating bone accumulations created by spotted hyenas and leopards: harnessing actualistic data from east and Southern Africa. J Taphon 8:155–179Google Scholar
  26. Domínguez-Rodrigo M, Rayne Pickering T, Semaw S, Rogers MJ (2005) Cutmarked bones from Pliocene archaeological sites at Gona, Afar, Ethiopia: implications for the function of the world’s oldest stone tools. J Hum Evol 48:109–121. CrossRefGoogle Scholar
  27. Domínguez-Rodrigo M, Saladié P, Cáceres I, Huguet R, Yravedra J, Rodríguez-Hidalgo A, Martín P, Pineda A, Marín J, Gené C, Aramendi J, Cobo-Sánchez L (2017) Use and abuse of cut mark analyses: the Rorschach effect. J Archaeol Sci 86:14–23. CrossRefGoogle Scholar
  28. Domínguez-Rodrigo M, Yravedra J (2009) Why are cut mark frequencies in archaeofaunal assemblages so variable? A multivariate analysis. J Archaeol Sci 36:884–894. CrossRefGoogle Scholar
  29. Domínguez-Rodrigo M, Yravedra J, Organista E, Gidna A, Fourvel JB, Baquedano E (2015) A new methodological approach to the taphonomic study of paleontological and archaeological faunal assemblages: a preliminary case study from Olduvai Gorge (Tanzania). J Archaeol Sci 59:35–53. CrossRefGoogle Scholar
  30. Efron B (1979) Bootstrap methods: another look at the jackknife. Ann Stat 7:1–26. CrossRefGoogle Scholar
  31. Egeland CP, Domínguez-Rodrigo M, Barba R (2007) The hunting-versus-scavenging debate. In: Domínguez-Rodrigo M, Egido RB, Egeland CP (eds) Deconstructing Olduvai: a taphonomic study of the bed I sites. Springer Netherlands, Dordrecht, pp 11–22CrossRefGoogle Scholar
  32. Egeland CP, Domínguez-Rodrigo M, Pickering TR, Menter CG, Heaton JL (2018) Hominin skeletal part abundances and claims of deliberate disposal of corpses in the Middle Pleistocene. Proc Natl Acad Sci 115(18):4601–4606; 201718678. CrossRefGoogle Scholar
  33. Ferretti MP (2007) Evolution of bone-cracking adaptations in hyaenids (Mammalia, Carnivora). Swiss J Geosci 100:41–52. CrossRefGoogle Scholar
  34. Fosse P, Avery G, Fourvel JB, et al (2010) Los cubiles actuales de hiena: síntesis crítica de sus características tafonómicas a partir de la excavación de nuevos yacimientos (República de Djibuti, Africa del Sur) y la información publicada. Zona Arqueol 108–117Google Scholar
  35. Fosse P, Selva N, Fourvel JB et al (2012) Bone modification by modern wolf (canis lupus): a taphonomy study from their natural feeding places. J Taphon 10:197–217Google Scholar
  36. Fourvel J-B, Fosse P, Avery G (2015) Spotted, striped or brown? Taphonomic studies at dens of extant hyaenas in eastern and Southern Africa. Quat Int 369:38–50. CrossRefGoogle Scholar
  37. Galán AB, Rodríguez M, de Juana S, Domínguez-Rodrigo M (2009) A new experimental study on percussion marks and notches and their bearing on the interpretation of hammerstone-broken faunal assemblages. J Archaeol Sci 36:776–784CrossRefGoogle Scholar
  38. Gidna A, Domínguez-Rodrigo M, Pickering TR (2015) Patterns of bovid long limb bone modification created by wild and captive leopards and their relevance to the elaboration of referential frameworks for paleoanthropology. J Archaeol Sci Rep 2:302–309. Google Scholar
  39. Gidna A, Yravedra J, Domínguez-Rodrigo M (2013) A cautionary note on the use of captive carnivores to model wild predator behavior: a comparison of bone modification patterns on long bones by captive and wild lions. J Archaeol Sci 40:1903–1910. CrossRefGoogle Scholar
  40. Gifford-Gonzalez DP (1989) Ethnographic analogues for interpreting modified bones: some cases from East Africa. In: Bonnichsen R, Sorg MH (eds) Bone modifications. Center for the Study of the First Americans, Institute for Quaternary Studies, University of Maine, Orono, pp 179–246Google Scholar
  41. Green NE, Swiontkowski MF (2000) Traumatismo esquelético en niños. Ed. Médica PanamericanaGoogle Scholar
  42. Harris JA, Marean CW, Ogle K, Thompson J (2017) The trajectory of bone surface modification studies in paleoanthropology and a new Bayesian solution to the identification controversy. J Hum Evol 110:69–81. CrossRefGoogle Scholar
  43. Haynes G (1983a) Frequencies of spiral and green-bone fractures on ungulate limb bones in modern surface assemblages. Am Antiq 48:102–114. CrossRefGoogle Scholar
  44. Haynes G (1983b) A guide for differentiating mammalian carnivore taxa responsible for gnaw damage to herbivore limb bones. Paleobiology 9:164–172CrossRefGoogle Scholar
  45. James EC, Thompson JC (2015) On bad terms: problems and solutions within zooarchaeological bone surface modification studies. Environ Archaeol 20:89–103. CrossRefGoogle Scholar
  46. Johnson E (1985) 5—current developments in bone technology. In: Schiffer MB (ed) Advances in archaeological method and theory. Academic, San Diego, pp 157–235CrossRefGoogle Scholar
  47. Karr LP, Outram AK (2012) Tracking changes in bone fracture morphology over time: environment, taphonomy, and the archaeological record. J Archaeol Sci 39:555–559. CrossRefGoogle Scholar
  48. Kuhn M (2017) Caret: classification and regression trainingGoogle Scholar
  49. Kuhn M, Johnson K (2013) Applied predictive modeling. Springer, New YorkCrossRefGoogle Scholar
  50. Lantz B (2013) Machine learning with R. Packt, BirminghamGoogle Scholar
  51. Lemon J (2006) Plotrix: a package in the red light district of R. R-News 6(4):8–12Google Scholar
  52. Lyman RL (1994) Vertebrate Taphonomy. Cambridge University PressGoogle Scholar
  53. Moclán A, Domínguez-Rodrigo M (2018) An experimental study of the patterned nature of anthropogenic bone breakage and its impact on bone surface modification frequencies. J Archaeol Sci 96:1–13. CrossRefGoogle Scholar
  54. Morlan RE (1984) Toward the definition of criteria for the recognition of artificial bone alterations. Quat Res 22:160–171. CrossRefGoogle Scholar
  55. Myers TP, Voorhies MR, Corner RG (1980) Spiral fractures and bone pseudotools at paleontological sites. Am Antiq 45:483–490. CrossRefGoogle Scholar
  56. Organista E, Pernas-Hernández M, Gidna A, Yravedra J, Domínguez-Rodrigo M (2016) An experimental lion-to-hammerstone model and its relevance to understand hominin-carnivore interactions in the archeological record. J Archaeol Sci 66:69–77. CrossRefGoogle Scholar
  57. Outram AK (2001) A new approach to identifying bone marrow and grease exploitation: why the “indeterminate” fragments should not be ignored. J Archaeol Sci 28:401–410CrossRefGoogle Scholar
  58. Outram AK, Knüsel CJ, Knight S, Harding AF (2005) Understanding complex fragmented assemblages of human and animal remains: a fully integrated approach. J Archaeol Sci 32:1699–1710. CrossRefGoogle Scholar
  59. Pante MC, Blumenschine RJ, Capaldo SD, Scott RS (2012) Validation of bone surface modification models for inferring fossil hominin and carnivore feeding interactions, with reapplication to FLK 22, Olduvai Gorge, Tanzania. J Hum Evol 63:395–407. CrossRefGoogle Scholar
  60. Parkinson JA (2018) Revisiting the hunting-versus-scavenging debate at FLK Zinj: a GIS spatial analysis of bone surface modifications produced by hominins and carnivores in the FLK 22 assemblage, Olduvai Gorge, Tanzania. Palaeogeogr Palaeoclimatol Palaeoecol 511:29–51. CrossRefGoogle Scholar
  61. Pickering TR (2002) Reconsideration of criteria for differentiating faunal assemblages accumulated by hyenas and hominids. Int J Osteoarchaeol 12:127–141. CrossRefGoogle Scholar
  62. Pickering TR, Domínguez-Rodrigo M, Egeland CP, Brain CK (2005) The contribution of limb bone fracture patterns to reconstructing early hominid behaviour at Swartkrans cave (South Africa): archaeological application of a new analytical method. Int J Osteoarchaeol 15:247–260. CrossRefGoogle Scholar
  63. Pickering TR, Egeland CP (2006) Experimental patterns of hammerstone percussion damage on bones: implications for inferences of carcass processing by humans. J Archaeol Sci 33:459–469. CrossRefGoogle Scholar
  64. Pineda A, Cáceres I, Saladié P, Huguet R, Morales JI, Rosas A, Vallverdú J (2019) Tumbling effects on bone surface modifications (BSM): an experimental application on archaeological deposits from the Barranc de la Boella site (Tarragona, Spain). J Archaeol Sci 102:35–47. CrossRefGoogle Scholar
  65. Pineda A, Saladié P (2018) The Middle Pleistocene site of Torralba (Soria, Spain): a taphonomic view of the Marquis of Cerralbo and Howell faunal collections. Archaeol Anthropol Sci.
  66. Pineda A, Saladié P, Huguet R, Cáceres I, Rosas A, Estalrrich A, García-Tabernero A, Vallverdú J (2017) Changing competition dynamics among predators at the late Early Pleistocene site Barranc de la Boella (Tarragona, Spain). Palaeogeogr Palaeoclimatol Palaeoecol 477:10–26. CrossRefGoogle Scholar
  67. Pokines JT, Kerbis Peterhans JC (2007) Spotted hyena (Crocuta crocuta) den use and taphonomy in the Masai Mara National Reserve, Kenya. J Archaeol Sci 34:1914–1931. CrossRefGoogle Scholar
  68. Prendergast ME, Domínguez-Rodrigo M (2008) Taphonomic analyses of a hyena den and a natural-death assemblage near Lake Eyasi (Tanzania). J Taphon 6:301–335Google Scholar
  69. Ringrose TJ (2013) cabootcrs: bootstrap confidence regions for correspondence analysisGoogle Scholar
  70. Sala N, Arsuaga JL, Haynes G (2014) Taphonomic comparison of bone modifications caused by wild and captive wolves (Canis lupus). Quat Int 330:126–135. CrossRefGoogle Scholar
  71. Selvaggio MM (1994) Identifying the timing and sequence of hominid and carnivore involvement with Plio-Pleistocene bone assemblages from carnivore tooth marks and stone-tool butchery marks on bone surfaces. Thesis Doctoral, Rutgers UniversityGoogle Scholar
  72. Stiner MC (2004) Comparative ecology and taphonomy of spotted hyenas, humans, and wolves in Pleistocene Italy. Rev Paléobiol 23:771–785Google Scholar
  73. Tseng ZJ, Binder WJ (2010) Mandibular biomechanics of Crocuta crocuta, Canis lupus, and the late Miocene Dinocrocuta gigantea (Carnivora, Mammalia). Zool J Linnean Soc 158:683–696. CrossRefGoogle Scholar
  74. Turner A, Antón M, Werdelin L (2008) Taxonomy and evolutionary patterns in the fossil Hyaenidae of Europe. Geobios 41:677–687. CrossRefGoogle Scholar
  75. Villa P, Mahieu E (1991) Breakage patterns of human long bones. J Hum Evol 21:27–48. CrossRefGoogle Scholar
  76. Wauthoz B, Dorning KJ, Hérissé AL (2003) Crassiangulina variacornuta sp. nov. from the late Llandovery and its bearing on Silurian and Devonian acritarch taxonomy. Bull Société Géologique Fr 174:67–81. CrossRefGoogle Scholar
  77. Werdelin L (1989) Constraint and adaptation in the bone-cracking canid Osteoborus (Mammalia: Canidae). Paleobiology 15:387–401CrossRefGoogle Scholar
  78. White EM, Hannus LA (1983) Chemical weathering of bone in archaeological soils. Am Antiq 48:316–322. CrossRefGoogle Scholar
  79. Willey P, Snyder LM (1989) Canid modification of human remains: implications for time-since-death estimations. J Forensic Sci 34:894–901CrossRefGoogle Scholar
  80. Yravedra J, Domínguez-Rodrigo M, Santonja M, Rubio-Jara S, Panera J, Pérez-González A, Uribelarrea D, Egeland C, Mabulla AZP, Baquedano E (2016) The larger mammal palimpsest from TK (Thiongo Korongo), bed II, Olduvai Gorge, Tanzania. Quat Int 417:3–15. CrossRefGoogle Scholar
  81. Yravedra J, Lagos L, Bárcena F (2011) A taphonomic study of wild wolf (Canis lupus) modification of horse bones in northwestern Spain. J Taphon 9:37–65Google Scholar
  82. Yravedra J, Lagos L, Bárcena F (2012) The wild wolf (Canis lupus) as a dispersal agent of animal carcasses in northwestern Spain. J Taphon 10:219–238Google Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2019

Authors and Affiliations

  1. 1.Centro Nacional de Investigación sobre la Evolución Humana (CENIEH)BurgosSpain
  2. 2.Escuela Interuniversitaria de Posgrado en Evolución HumanaUniversidad de BurgosBurgosSpain
  3. 3.Institute of Evolution in Africa (IDEA)University of Alcalá de HenaresMadridSpain
  4. 4.Departamento de PrehistoriaUniversidad Complutense de MadridMadridSpain

Personalised recommendations