# Carnivore tooth marks and stone tool butchery marks on scavenged bones: archaeological implications

> What is left of a carcass when the lions, leopards, hyenas and jackals walk away? In the Serengeti, a researcher watched carnivores feed, then scavenged the leftover limbs with stone tools. Bones left by lone feeders still carried flesh, while those fought over by big mixed groups held little but marrow, and the tooth marks, cut marks and hammerstone marks on the bones record that difference.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/1994-selvaggio-carnivore-tooth-marks-stone/

## Citation and access

- **Authors:** Marie M. Selvaggio
- **Year published:** 1994
- **Journal or source:** Journal of Human Evolution
- **Full citation:** Selvaggio, M. M. 1994. Carnivore tooth marks and stone tool butchery marks on scavenged bones: archaeological implications. Journal of Human Evolution 27(1–3): 215–228.
- **DOI:** 10.1006/jhev.1994.1043
- **Access status:** Publisher link only
- **Where to find it:** https://doi.org/10.1006/jhev.1994.1043

## Tags

- **Species:** Mammals > Cheetah; Mammals > Jackal; Mammals > Leopard; Mammals > Lion; Mammals > Spotted hyena
- **Region:** Africa > Tanzania
- **Tracking:** Bones; Feeding signs; Tracking; Tracks and signs identification
- **Behaviour and ecology:** Behaviour; Competition; Feeding; Scavenging
- **Methods and technology:** Bone damage analysis; Descriptive statistics; Direct observation
- **Data type:** New field data
- **Observer qualification:** Not reported
- **Discipline:** Behaviour and ecology; Palaeoanthropology; Taphonomy
- **Year published:** 1994

## Detailed keywords

- **Species:** Studied (carnivores watched eating the limbs, whose tooth marks were recorded): lions, leopards, cheetahs, spotted hyenas and jackals. No scientific names are given, and the jackal species is not named. Carcasses: 32 prey animals; the species are not named, only sizes (“small gazelle-sized prey” and prey larger than gazelles). Discussed: Plio-Pleistocene hominids (early human relatives) as possible hunters or scavengers, and wildebeest-sized prey.
- **Location and habitat:** Serengeti National Park and Ngorongoro Conservation Area, northern Tanzania: tropical savanna-woodland with migratory and resident large herbivores and carnivores. Archaeological site discussed: FLK Zinjanthropus (FLK 22, the “Zinj floor”), Olduvai Gorge, Tanzania.
- **Topics:** hunting versus scavenging by early hominids; early versus late access to carcasses; carnivore feeding behaviour and carcass abandonment; competition at carcasses and feeding group size; carnivore tooth marks; stone tool cut marks; hammerstone percussion marks; marrow versus flesh scraps left on long bones; experimental butchery with basalt and quartzite; long bone portions (complete bones, epiphyses, near-epiphyses, midshafts); identifying marks with a hand lens; experimental archaeology; bone modification; taphonomy.
- **Statistics or models:** Descriptive statistics: mean assemblage percent with ±1 standard deviation; ratios of cut marks to tooth marks and of percussion marks to cut marks, by feeding group size; simple conditional probabilities of group size given a ratio (Tables 3b and 4b). Inconsistencies in the paper: (1) Table 4b gives a probability of 0.00 for large groups at a percussion-to-cut-mark ratio of 5:1 or more, but by Table 4a only large groups reached that ratio (75% of their assemblages), so the value should be 1.00; (2) Table 4b gives 0.24 for large groups at 2:1, while the stated method gives about 0.25; (3) the text says the highest cut-to-tooth ratios, “greater than 3 to 1”, belong to solitary consumers, while Table 3a’s top column is “4:1 or more” and Figure 3 shows solitary assemblages at 7:1 and 10:1; the summary gives “7:1 or more”.
- **Data type:** New field data: 7 months of observations of free-ranging carnivores in 1989 to 1990, with experimental butchery of the long bones collected after the carnivores left. Sample: 168 long bones from 32 carcasses, broken into 584 specimens of 2 cm or longer.

## Synopsis

### 1. What they did

Did early hominids (early human relatives) of the Plio-Pleistocene get meat by hunting, or by scavenging what carnivores left behind? Archaeologists try to answer this from the marks on fossil bones: carnivore tooth marks, cut marks from stone tools, and percussion marks where a hammerstone broke a bone open for marrow. Earlier experiments had people butcher bones first and let carnivores chew them afterwards. This study modelled the reverse: carnivores eat first, and a scavenger with stone tools comes later.

- Watching carnivores eat. Over 7 months in 1989 to 1990, in Serengeti National Park and the Ngorongoro Conservation Area, Tanzania, the author watched free-ranging lions, leopards, cheetahs, spotted hyenas and jackals eat carcasses, from a four-wheel-drive vehicle placed to see well with little disturbance. Every limb in the sample was seen being eaten by carnivores; the hunt itself was seen for 19 of the carcasses.

- Collecting the bones. Once the carnivores had left (and were about 0.5 km away), she collected the long bones that still held marrow and sometimes scraps of flesh.

- Butchering them with stone. Using basalt and quartzite, stones common at early archaeological sites, she (and, for about half the experiments, a Tanzanian butcher who did not know their purpose) cut the limbs apart, removed leftover flesh, skin and tendons, and broke the bones with a hammerstone to reach the marrow.

- Counting the marks. The cleaned bones were examined under a 16 power hand lens in strong, low-angle light, using published criteria to tell tooth marks from butchery marks. Only pieces 2 cm or longer were counted.

- Comparing feeding groups. The 32 carcasses were split by how many carnivores had fed: solitary animals (15 carcasses), small groups of two to four (13), and large groups of nine or 11 (4).

### 2. What they found

Carnivores at the carcass

- Solitary feeders had no competition and often left limbs barely pulled apart, with skin, tendons and flesh scraps still on them.

- Small groups, often closely related animals such as mothers with offspring or members of one lion pride, showed only moderate aggression and also left flesh scraps.

- Every large group was a mix of at least two kinds of carnivore (cats, spotted hyenas and/or jackals). Competition was most intense here: limbs were pulled apart and the bone ends were often heavily damaged or gnawed away. Marrow was usually the only food left.

- Large-group carcasses were rare to scavenge during the study (only 4).

Marks on the bones (Table 2)

- Across all 584 pieces, more carried tooth marks (65% on average) than percussion marks (46%) or cut marks (34%).

- Bone ends (epiphyses) were the most often tooth-marked (75%) and cut-marked (37%), as they are where muscles attach and joints are cut apart. Midshafts were the least often marked by teeth (47%) or tools (17%).

- On the 35 bones left whole (not broken for marrow), cut marks (67%) were more common than tooth marks (49%), because these came from solitary and small-group carcasses where people still had to cut the limbs apart and remove flesh.

- Standard deviations were large, showing great variation between carcasses.

Ratios that reflect the number of carnivores

- Cut marks to tooth marks: ratios of 2:1 or higher came only from bones left by solitary animals and small groups; the highest ratios came only from solitary feeders. Bones left by large groups had the lowest ratios (half at 0:1, half at 1:1).

- Percussion marks to cut marks: 75% of large-group assemblages had 5:1 or more, while every solitary and small-group assemblage was below 5:1.

### 3. What they concluded

- How many carnivores fed on a carcass, and how fiercely they competed, decides whether flesh scraps or only marrow are left for a scavenger. This leaves predictable, recognisable marks on the bones.

- Where fossil bones are well preserved, the balance of tooth marks, cut marks and percussion marks could be used to estimate how many carnivores fed first, and so whether a hominid scavenger got flesh or mainly marrow.

- If hominids had come first and carnivores later, cut-to-tooth ratios should be higher than in this study, and percussion-to-cut ratios should stay in the lowest categories.

- Midshaft pieces are best for this, because carnivores scavenging later tend to ignore marrow-emptied midshaft fragments.

- At FLK Zinjanthropus (Olduvai Gorge), the same bones have been used to argue both that hominids came first and that carnivores did. The author argues that both views overlook the range of carnivore feeding behaviour, and that observations of living carnivores plus experiments can help settle the question.

### 4. How tracking was or was not used

No tracks or trailing were used. The study links directly observed carnivore feeding to one kind of sign: tooth marks on bones, read alongside cut and percussion marks. The author identified the marks herself using published criteria and a hand lens. Seven pieces from young animals, whose bone surface made marks hard to identify, were left out.

### 6. Other technologies, skills or innovations

- Experimental butchery with basalt and quartzite stone tools and hammerstones, with a standard routine: no scraping of the bone, first blow near the top end of each bone, then blows along the length until the marrow showed.

- A 16 power hand lens under strong, low-angle light for identifying marks.

- Comparing ratios of mark types rather than counts alone, to control for the extra marks created when bones are smashed.

### 8. Implications for tracking (Original Wisdom's observations)

_These are Original Wisdom's observations, not the authors'._

- The study shows that what carnivores leave behind depends on who fed and how many: a lone feeder leaves limbs barely pulled apart with flesh still on them, while a mixed group of cats, hyenas and jackals leaves limbs separated and bone ends gnawed away. These are feeding signs a tracker can read at a carcass.

## Cite this summary

Lawrence, K. 2026. Summary of: Carnivore tooth marks and stone tool butchery marks on scavenged bones: archaeological implications (Selvaggio, 1994). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/1994-selvaggio-carnivore-tooth-marks-stone/