The utility of osteon shape and circularity for differentiating human and non‐human Haversian bone

Abstract

Distinguishing human from non-human bone fragments is usually accomplished by observation of gross morphology. When macroscopic analysis is insufficient, histological approaches can be applied. Microscopic features, like plexiform bone or osteon banding, are characteristic of non-humans. In the absence of such features, distinguishing Haversian bone as either human or non-human proves problematic. This study proposes a histomorphometric approach for classifying species from Haversian bone. Two variables, osteon area (On.Ar.) and circularity (On.Cr.), are examined. Measurements were collected from three species (deer, dog, human) represented by various skeletal elements; only ribs were available for humans (ribs: deer n = 6, dog n = 6, human n = 26; humeri: deer n = 6, dog n = 6; femora: deer n = 6, dog n = 6). Qualitative analysis comparing human to non-human On.Ar. demonstrated that human ribs have larger mean On.Ar. (0.036 mm(2)) than non-human ribs (deer = 0.017 mm(2) , dog = 0.013 mm(2)). On.Cr. in the ribs showed minor differences between species (deer = 0.877; dog = 0.885; human = 0.898). Results demonstrated no significant difference across long bone quadrants in long bones. Discriminant analyses run on the means for each sample demonstrated overlap in deer and dog samples, clustering the non-human and human groups apart from each other. Mean On.Cr. proved a poor criterion (ribs only: 76.3%, pooled elements: 66.1%), while mean On.Ar. proved useful in identifying human from non-human samples (ribs only: 92.1%, pooled elements: 93.5%). When variables were combined, accuracy increased to 100% correct classification for rib data and 98.4% when considering data from all elements. These results indicate that On.Ar. and On.Cr. are valuable histomorphometric tools for distinguishing human from non-human Haversian bone.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 59 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.