# Herders’ ecological knowledge and carnivore predation on livestock investigations in Makgadikgadi and Nxai national parks, Botswana

> In Botswana, herders often carry out the first check when a lion, leopard or wild dog kills their livestock, and compensation depends on naming the right predator. Researchers found that herders’ reports of predation matched official records well, and they recognised carnivores easily from photos. Identifying the same carnivores from photos of their tracks and kills proved much harder.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/2017-rutina-mogwera-herders-ecological-knowledge-carnivore/

## Citation and access

- **Authors:** Lucas P. Rutina, Kefentse M. Mogwera, Elford Seonyatseng, Charles Mpofu and Ditso Ntloyathuto
- **Year published:** 2017
- **Journal or source:** Koedoe
- **Full citation:** Rutina, L. P., Mogwera, K. M., Seonyatseng, E., Mpofu, C. and Ntloyathuto, D. 2017. Herders’ ecological knowledge and carnivore predation on livestock investigations in Makgadikgadi and Nxai national parks, Botswana. Koedoe 59(2): a1389.
- **DOI:** 10.4102/koedoe.v59i2.1389
- **Access status:** Open access (CC BY 4.0)
- **PDF:** https://www.originalwisdom.com/wp-content/uploads/2026/11/2017_Rutina_and_Mogwera_etal_herders-ecological-knowledge-and-carnivore-predation-on-livestock-investigations.pdf

## Tags

- **Species:** Mammals > African wild dog; Mammals > Cheetah; Humans; Mammals > Leopard; Mammals > Lion; Mammals > Spotted hyena
- **Region:** Africa > Botswana
- **Tracking:** Feeding signs; Track photographs; Tracking; Tracks; Tracks and signs identification
- **Behaviour and ecology:** Behaviour; Predation
- **Methods and technology:** Chi-square test; Global Positioning System (GPS); Interviews; Observer reliability; Survey instruments
- **Data type:** Existing data; New field data
- **Observer qualification:** Other experienced trackers
- **Discipline:** Human dimensions; Human–wildlife conflict; Social science; Wildlife tracking
- **Year published:** 2017

## Detailed keywords

- **Species:** Studied: humans (livestock herders, both owners and employed herders), and their identification of five large carnivores: lion (Panthera leo), leopard (Panthera pardus), African wild dog (Lycaon pictus), cheetah (Acinonyx jubatus) and spotted hyena (Crocuta crocuta). Predation records and interviews covered lion, leopard and wild dog; herders also reported spotted hyena damage, which is not compensated. In the study area but not studied: zebra (Equus quagga), springbok (Antidorcas marsupialis), blue wildebeest (Connochaetes taurinus) and gemsbok (Oryx gazella).
- **Location and habitat:** Communal area CT 7 (about 6543 km²), east of Makgadikgadi Pans National Park (NG 52 and CT 9) and Nxai Pan National Park, between wildlife management areas CT 1 and CT 2 (north) and CT 11 (south), northern Botswana. Workshop sections: Gweta north, Gweta south and Nogatshaa. 63 cattle posts, 9 within 10 km of the parks (“near”) and 54 beyond 10 km (“far”). Semi-arid, 250 to 450 mm rainfall, mostly November to April; trees include baobab (Adansonia digitata), leadwood (Combretum imberbe), silver terminalia (Terminalia sericea), mopane (Colophospermum mopane) and real fan palm (Hyphaene petersiana). Part of the Kavango-Zambezi Transfrontier Conservation Area (KAZA TFCA).
- **Topics:** local ecological knowledge (LEK) of herders; livestock predation (depredation) by large carnivores; human-wildlife conflict; compensation investigations by the Department of Wildlife and National Parks (DWNP) Problem Animal Control (PAC) unit; identifying carnivores by sight, tracks and kill patterns (bite marks, body parts eaten, carcass location); picture-card identification tests; participatory mapping of predation risk at cattle posts; distance from protected areas; community-based natural resource management; decentralising conflict management. Inconsistencies in the paper: (1) the abstract says there was a significant association between reported and perceived predation for all carnivores at all distances, while the Results say there was no significant association (chi-square, p > 0.05 for all tests); (2) the Kruskal-Wallis result for lion risk is given as significant for cattle posts far from the parks, but the post hoc test and the next sentence say the difference was significant only near the parks; (3) the summary percentages for correct identification (visual near 76%, far 75%; tracks near 35%, far 63%; kill patterns near 35%, far 34%) cannot be matched to the per-species bars in Figure 5 (for example, tracks near the parks range from about 11% to 39% by species); (4) the text gives Nxai Pan National Park as NG 50, the map as NG 48; (5) the text cites Figure 4 for spotted hyena damage, but Figure 4 shows only lion, leopard and wild dog; (6) 18 of 40 self-employed herders is given as 46% (it is 45%).
- **Statistics or models:** Chi-square (χ²) tests; Kruskal-Wallis test with Dunn-Bonferroni post hoc test; conditional probability that both data sources place a cattle post in the same risk category; percentages of correct identifications.
- **Data type:** Existing data: DWNP Problem Animal Control records, 2008 to 2012. New field data: semi-structured questionnaire interviews with 62 herders, a mapping workshop and picture-card identification test with 30 herders, and GPS locations of cattle posts. Dates of interviews and workshop not given. Also uses DWNP investigation diaries.

## Synopsis

### 1. What they did

Botswana pays farmers compensation when wildlife kills their livestock, but only after an investigation confirms which species did it. Because the Department of Wildlife and National Parks (DWNP) is short of staff and resources, herders often carry out the first investigation themselves. They rely on three indicators: seeing the carnivore at the kill, the predator’s tracks, and the kill pattern on the carcass (bite marks, which parts were eaten and where the carcass lies). The authors set out to test how well herders’ local ecological knowledge (LEK) holds up.

The study took place in a communal farming area of about 6543 km² east of Makgadikgadi Pans and Nxai Pan national parks, with 63 cattle posts.

- Official records. They took five years (2008 to 2012) of livestock predation incidents investigated and confirmed by the DWNP Problem Animal Control (PAC) unit for lion, leopard and African wild dog.

- Interviews. They interviewed 62 herders, randomly chosen from farmers who had reported predation: 11 from cattle posts within 10 km of the parks and 51 from farther away. The herders were asked how long they had lived in the area, how many livestock they had lost to each carnivore, and how they saw the predation risk.

- Mapping workshop. Thirty of the herders (10 women and 20 men), chosen for their experience and time in the area, mapped each cattle post in their section as low, medium or high risk for each carnivore.

- Identification test. The same 30 herders were shown picture cards and asked to identify cheetah, leopard, lion, spotted hyena and wild dog from photos of the animal, of their tracks, and of kills.

They then compared the herders’ reports with the official records.

### 2. What they found

Herders’ reports versus official records

- Both sources showed that most cattle posts had low predation: more than 60% for lion, 75% for leopard and 80% for wild dog.

- Near the parks, official lion incidents were higher than the herders’ reports; farther away there was no difference. For leopard and wild dog, herders reported more incidents than the official records at all distances.

- Herders reported that spotted hyena caused more damage than the three compensated species, especially far from the parks. There are no official records for hyena, because hyena damage is not compensated.

- Both sources agreed that lion predation falls with distance from the parks and that leopard incidents were spread evenly. They disagreed on wild dog: official records showed an even spread, herders said wild dog predation increased closer to the parks.

- The chance that both sources placed a cattle post in the same risk category was 70% (far) and 56% (near) for lion, 67% and 56% for leopard, and 63% and 78% for wild dog.

Identification from picture cards

- Herders identified the large carnivores well from photos of the animals: 76% correct near the parks and 75% far away.

- They found tracks and kill patterns much harder. Tracks: 35% correct near the parks and 63% far away. Kill patterns: 35% near and 34% far.

- Results varied by species (Figure 5). For example, lion tracks were identified correctly by few herders near the parks but by most herders far away.

- In contrast, DWNP investigation diaries showed that when herders used their own knowledge, mostly tracks and kill patterns, to name the carnivore responsible, DWNP confirmed their identification in 82% of lion cases, 67% of wild dog cases and 100% of leopard cases.

The paper reports some results inconsistently (see Detailed keywords).

### 3. What they concluded

- Herders’ knowledge and the official records complemented each other, both in reporting predation incidents and in ranking cattle posts by risk, so local communities should be involved in predation investigations and management.

- Herders could recognise the carnivores themselves but had difficulty with tracks and kill patterns. The authors suggest this may come from current herding practice: herders are not paid for all herding work, so they take livestock to water and bring them back to the kraal (boma) in the afternoon. They also suggest that poor pictures on the cards may partly explain the low scores, since the diaries show herders do identify predators from tracks and kills.

- They recommend more education for herders on identifying carnivore tracks and kill behaviour.

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

The study tested tracks and signs identification rather than using it to collect data. Herders were shown photographs of carnivore tracks (one example is a single track in sand) and of carcasses showing kill patterns, and asked which species made them. No trailing was involved. The paper also notes that tracks are one of the three indicators used in official compensation investigations, and that DWNP diaries show herders mostly identify predators from tracks and kill patterns.

### 5. Tracker or observer certification

The herders are the subject of the study. They were chosen for the workshop for their experience and time in the area, and their identification skill was tested in the study with picture cards. No certification is reported. The authors who designed and marked the test are not described as trackers, and the paper does not say how the correct answers on the track and kill cards were confirmed.

### 6. Other technologies, skills or innovations

- Picture cards of the animal, a track and a kill pattern. The kill-pattern photo came from the Cheetah Conservation Fund in Namibia and shows a carcass with marked areas.

- Participatory risk mapping on paper maps, with each cattle post scored as low (0 to 3 incidents a month), medium (4 to 6) or high (more than 6) risk for each carnivore.

- Handheld Global Positioning System (GPS) device to record cattle post locations, because official records gave no precise locations.

### 7. Implications for conservation (from the paper)

- Decentralising human-wildlife conflict management to local communities through existing local structures (district conservation committees, village development committees and village farmer committees) would add to DWNP’s limited resources and give communities a role in wildlife management.

- Training communities to collect monitoring data, for example with a Management-Oriented Monitoring System (MOMS), would improve information on conflict between people and carnivores.

- Mitigation needs to be specific to each area and species, rather than “one size fits all”.

- Involving local communities in managing conflict could help the Kavango-Zambezi Transfrontier Conservation Area (KAZA TFCA) meet its aims.

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

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

- Herders who live with carnivores every day recognised the animals well but scored much lower on their tracks and kill patterns in this test. Living alongside wildlife does not automatically bring reliable tracks and signs identification; skill needs to be tested rather than assumed.

- Identifying tracks from a single photograph is harder than reading tracks on the ground, where a tracker can follow the trail and see more signs. The authors themselves suggest poor pictures may have lowered the scores, and the diaries show high agreement with official confirmations. A field test may give a different picture of the herders’ skill.

- When compensation depends on naming the predator, the accuracy of tracks and signs identification has real consequences for farmers and for carnivores.

## Cite this summary

Lawrence, K. 2026. Summary of: Herders’ ecological knowledge and carnivore predation on livestock investigations in Makgadikgadi and Nxai national parks, Botswana (Rutina and Mogwera et al., 2017). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/2017-rutina-mogwera-herders-ecological-knowledge-carnivore/