Tracking and the interpretation of spoor: a scientifically sound method in ecology

1. What they did

Biologist P. E. Stander and four experienced Ju/’Hoan San hunters set out to test whether tracking is reliable enough to count as a scientific method, and to understand the techniques the trackers use. The trackers are co-authors.

• Stander observed wild animals directly in Kaudom Game Reserve and Bushmanland, Namibia, recording species, age, sex, group composition, position and behaviour.
• The trackers were then taken “blind” to the general area within 24 hours. They were told only the broad group of animal (for example “carnivore” if a lion had been seen), and asked to find the spoor and reconstruct what had happened.
• The four hunters worked as one team, reaching group decisions after critical discussion, as is usual in Ju/’Hoan society. Each answer was scored correct or incorrect against Stander’s observations; extra information they volunteered was ignored.
• Stander also measured spoor size and shape, and the spacing of feet in different gaits, to understand how trackers might tell animals and actions apart.

The authors describe the paper as addressing simple and systematic tracking (inductive reasoning), and not speculative tracking.

2. What they found

Overall accuracy

The team was correct in 557 of 569 cases (97.9%), with a standard error of only 0.1% of the mean. The few mistakes were trivial and did not affect the results statistically.

Identifying spoor

• Species: all 147 cases correct (lion, leopard, cheetah, spotted hyena, wild dog, steenbok, duiker, kudu, hartebeest, wildebeest, gemsbok and roan).
• Age: 64 of 69 correct. Leopard and cheetah were always right; with lions, sub-adults were sometimes called adults.
• Sex: 67 of 69 correct. Lion was always right; there was one mistake each for leopard and cheetah.
• Known individuals: 30 of 32 correct (93.8%): lion 11 of 12, leopard 16 of 17, cheetah 3 of 3. On average the team was choosing among at least about four known individuals likely to be in the area (range 2 to 11). The trackers said they recognized individuals by subtle differences in spoor size and shape, and by each animal’s way of walking.

Reconstructing behaviour

• The team correctly interpreted 249 of 252 behaviours (98.8%): all carnivore activities (lying, walking, stalking, crouching, running, capturing and killing prey, dragging carcasses and feeding); all prey details (species, age and sex, number, position and movement); whether carnivores had scavenged rather than killed; which other carnivores had visited a carcass; and the cause of death of four ungulates that died of anthrax.
• The only errors were in estimating when an event happened: 20 of 23 correct, using four daytime and two night-time periods.
• All 30 event summaries were correct.
• Example: an adult zebra died of anthrax. Over the next two days lions fed on it, followed by spotted hyenas and vultures. Three days after the death, the team followed the zebra’s trail and described how it had walked unsteadily, fallen, got up and walked on to where it died. Although they were not familiar with anthrax, they said it had been sick and died. They worked out that lions fed on it but did not kill it, that spotted hyenas and then vultures came after, and correctly estimated when it happened.

Spoor measurements

• Species differed in spoor size and in the proportions of front and back feet. Leopard and brown hyena spoor were symmetrical; some spotted hyenas and lions had slightly elongated spoor.
• Adult male lion spoor was significantly larger than adult female spoor. Sub-adult female spoor was smaller than sub-adult male spoor, and similar to adult females. Sub-adult males fell between adult females and adult males.
• Differences between individuals could not be shown statistically, because soil conditions varied so much. For well-measured individuals, the variation within one animal’s spoor was as large as the variation across the whole species.
• The trackers recognized and allowed for this soil variation, which the measurements could not.

Gaits

• Two adult male lions walking side by side at the same speed had significantly different strides, paces and straddles, supporting the trackers’ view that each animal has its own way of walking.
• Walking showed fairly even spacing between feet. Stalking lions and leopards took longer strides and placed the back foot in the front foot’s print (perfect register), and the straddle widened. Trotting leopards lengthened the pace and narrowed the straddle. A galloping lion used a rotary gallop, leading with the left front leg.

3. What they concluded

• Tracking by skilled Ju/’Hoan trackers is a reliable and accurate method for studying wildlife, including the behaviour of nocturnal carnivores in wooded habitat, where direct observation is difficult.
• Tracking has the advantages of being completely non-intrusive and possible in most habitats a day or two after an event. The authors used it to study lion and leopard hunting behaviour in the same area.
• The size of spoor alone should be used with caution to identify individuals, because soil affects measurements. They note that tiger census methods based on spoor size do not appear to allow for soil and terrain.
• The skills and knowledge of local people are often ignored, and local people are relegated to labouring roles. Indigenous perception of nature may be based on scientifically sound thinking.
• Collaboration between Western science and traditional knowledge could greatly benefit wildlife ecology and conservation.

4. How tracking was or was not used

Tracking was the subject of the study. Four Ju/’Hoan hunters, working as one team, identified spoor by species, age, sex and individual; followed trails; and reconstructed carnivore hunts, prey behaviour, scavenging, causes of death and the timing of events. Their answers were checked against direct observations of wild animals made by Stander before each test. Spoor and gait measurements were taken to understand the cues trackers use.

5. Tracker or observer certification

Not reported. The trackers are described as four experienced Ju/’Hoan hunters; no formal evaluation or certification is mentioned (the study predates the CyberTracker certification’s spread). Here, the study itself serves as a test of their accuracy.

6. Other technologies, skills or innovations

• Direct observation of wild animals was used as the known answer against which tracking was tested.
• A sampling-intensity method was used to decide how many spoor measurements were needed for a reliable average.
• Soil depth and the angle of the print walls were recorded, so that only clear prints in firmer soil were analysed.

7. Implications for conservation (from the paper)

• Tracking allows study of nocturnal and elusive carnivores in habitats where direct observation is limited, without disturbing the animals.
• Successful wildlife conservation depends largely on involving local inhabitants and incorporating their aspirations. Collaborating with local communities can uncover knowledge that has not yet been identified or documented.
• Census methods based on spoor size need to account for soil and terrain to be credible.