# How are people affecting the distribution of less migratory wildlife in the southern Kalahari of Botswana? A spatial analysis

> Why are gemsbok missing from so much of Botswana’s southern Kalahari? By counting tracks and dung at 43 points and comparing them with ground and aerial counts, researchers found that cattle change the vegetation only within about 10 km of their boreholes, yet gemsbok stay rare out to 40 km and beyond. They conclude that hunting, not competition with livestock, is pushing gemsbok away.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/1998-verlinden-perkins-how-people-affecting-distribution/

## Citation and access

- **Authors:** Alex Verlinden, Jeremy S. Perkins, Mark Murray and Gaseitsiwe Masunga
- **Year published:** 1998
- **Journal or source:** Journal of Arid Environments
- **Full citation:** Verlinden, A., Perkins, J. S., Murray, M. and Masunga, G. 1998. How are people affecting the distribution of less migratory wildlife in the southern Kalahari of Botswana? A spatial analysis. Journal of Arid Environments 38(1): 129–141.
- **DOI:** 10.1006/jare.1997.0298
- **Access status:** Publisher link only
- **Where to find it:** https://doi.org/10.1006/jare.1997.0298

## Tags

- **Species:** Mammals > Common duiker; Mammals > Gemsbok; Mammals > Scrub hare; Mammals > Springbok; Mammals > Springhare; Mammals > Steenbok
- **Region:** Africa > Botswana
- **Tracking:** Scat; Sign surveys; Track counts; Tracking; Tracks; Tracks and signs identification
- **Behaviour and ecology:** Behaviour; Habitat selection; Human disturbance
- **Methods and technology:** Aerial surveys; Distribution mapping; Geographic Information System (GIS); Global Positioning System (GPS); Moran’s I; Road transects
- **Data type:** Existing data; New field data
- **Observer qualification:** Not reported
- **Discipline:** Behaviour and ecology; Wildlife tracking
- **Year published:** 1998

## Detailed keywords

- **Species:** Studied: gemsbok (Oryx gazella), steenbok (Raphicerus campestris), duiker (Sylvicapra grimmia), springbok (Antidorcas marsupialis), springhare (Pedetes capensis) and scrub hare (Lepus saxatilis), together with livestock: cattle, donkeys, goats and sheep. Steenbok and duiker spoor were combined in the analysis. Present but too few for useful results: kudu (Tragelaphus strepsiceros). Spoor and ground counts were collected for many other game species, but the paper reports useful results only for the four common antelope.
- **Location and habitat:** South-western part of the Kalahari of Botswana, around Tsabong, along the border with South Africa: communal grazing areas, wildlife management areas (WMAs) and a portion of Gemsbok National Park (study area 20,294 km²). Nutrient-poor Kalahari sand, no permanent surface water, average rainfall about 300 mm at Tsabong with year-to-year variation as high as 45%. Livestock depend on boreholes (waterpoints, usually cattle-posts); pans attract springbok.
- **Topics:** human impact on wildlife distribution; livestock grazing versus hunting and poaching as explanations; borehole-based cattle keeping; livestock impact zones around waterpoints; bush encroachment and loss of grass cover; displacement of gemsbok by hunting pressure; subsistence hunting with horses, dogs and spears; hunting quotas; spoor (tracks and dung) indices; comparing spoor, ground counts and aerial surveys.
- **Statistics or models:** Spearman rank correlation; first-order asymptotic (step-function) models fitted by least squares; Kruskal-Wallis one-way analysis of variance by ranks; Moran’s I for spatial autocorrelation; second-order polynomial trend-surface regression (R² = 0.36); semi-variogram with a spherical model and ordinary kriging (GEO-EAS software); IDRISI Geographic Information System (GIS) distance images. Points to check in the paper: the spoor density classes overlap (“3–5 counts” and “5–10 counts” both include 5); the text says cattle tracks up to 20 km indicate effects on vegetation up to 20 km, while the impact zones say vegetation effects are not important at 10–20 km and the abstract gives 10 km; the variogram is reported with a nugget of 1000 and a sill of 7.50 (a sill smaller than the nugget is unusual and may be a typo); an R² of 0.36 is described as a “high” coefficient of determination.
- **Data type:** New field data: spoor and vegetation surveys at 43 points in September 1991 (peak of the dry season, during the hunting season); 16 ground counts from a vehicle, April 1991 to April 1992. Existing data: the average of six Department of Wildlife and National Parks aerial surveys, 1989 to 1993, and a 1990–1991 waterpoint survey.

## Synopsis

### 1. What they did

Since the 1950s, boreholes have opened the Kalahari to year-round cattle grazing, and wildlife numbers have fallen in many areas. Earlier work suggested that some less migratory wildlife avoid areas near people. The authors tested two possible reasons: (1) livestock changing the habitat, and (2) people hunting and poaching wildlife.

In the south-western Kalahari of Botswana, they compared four kinds of data against distance from the nearest livestock waterpoint:

- Spoor counts (September 1991). At 43 points, they walked a 100 m transect at right angles to a vehicle track, starting 50 m from the track, and counted wildlife and livestock tracks and dung. Wildlife counts were turned into a density score from 0 (not present) to 5 (very high, over 20); livestock spoor was often too dense to count and was scored directly on the same scale.

- Vegetation surveys in 20 x 50 m plots at the same points, estimating the cover of trees and shrubs, dwarf shrubs, grasses and bare ground.

- Ground counts. Sixteen counts from a vehicle driving at 40 km per hour between Tsabong and Gemsbok National Park (April 1991 to April 1992), giving the mean number of each species seen per 5 km.

- Aerial surveys. The average of six aerial counts from 1989 to 1993.

They mapped everything in a Geographic Information System (GIS) and used spatial statistics to allow for the fact that nearby points are not independent.

### 2. What they found

Livestock and vegetation

- Close to waterpoints, grass cover was low (about 15%) and woody plants had spread (bush encroachment). Grass cover rose to about 45% further away.

- Cattle spoor and cattle counts fell sharply with distance. Cattle were not seen beyond 13 km, though some cattle tracks were found up to 20 km. Goats and sheep stayed closer to water than cattle.

- The authors describe four zones: 0–1 km, very high impact, often bare ground; 1–5 km, high impact, low grass and bush encroachment; 5–10 km, low impact; 10–20 km, only occasional traces of cattle. The national park and the wildlife management areas showed no livestock effects; most effects were along the border with South Africa.

Wildlife

- Gemsbok were the clear exception. Their spoor and road counts rose steadily with distance from water, out to about 45–50 km, and they were likely very rare within 20 km. They were scarcer at 20–40 km than beyond 40 km, well past the zone where livestock change the vegetation. Only two gemsbok were seen inside the livestock impact zone in the aerial surveys. Trend-surface analysis and kriging of the aerial counts confirmed the same gradual gradient, which was hard to see by eye on the density maps.

- Steenbok and duiker spoor was slightly lower within about 8 km of waterpoints, but high densities still occurred well inside the zone of high livestock impact.

- Springbok were often common near waterpoints, apparently drawn to the shorter grass there, but also chose pans that have nothing to do with waterpoints.

- Springhare and scrub hare showed no link with distance to water.

- Kudu were too few to analyse.

### 3. What they concluded

- Habitat change caused by livestock appears to have little effect on most of these less migratory species.

- The gap between the livestock zone and high gemsbok densities points to hunting pressure, not competition with livestock, as the cause. The authors link this to a preference for gemsbok meat, hunting with horses, dogs and spears, and poaching. Many subsistence hunters said that hunting steenbok and duiker still takes more effort per kilogram of meat than hunting gemsbok.

- Springbok gathering in the same patches all year could make them easy to over-hunt.

- Overall, people using wildlife (hunting) matters more for the current distribution of common game in the southern Kalahari than animals avoiding areas changed by livestock.

- Aerial surveys combined with spoor data, waterpoint and vegetation surveys, and to a lesser extent ground counts, are useful for mapping species distributions in the Kalahari.

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

Spoor (tracks and dung) counts were one of the main data sources. Wildlife and livestock tracks and dung were counted along 100 m transects at 43 points and turned into density scores. Tracks and dung were identified with a reference collection the team gathered by following herds and identifying fresh tracks and dung, with Liebenberg’s (1990) field guide to the animal tracks of southern Africa as an extra reference. Steenbok and duiker spoor were combined rather than told apart. No trailing of individual animals is reported.

### 5. Tracker or observer certification

Not reported. The paper does not say who identified the spoor or how skilled they were. The acknowledgements thank field assistants Baipidi Pelekekae and Selefo Sabone, “especially with spoor identification”. The findings for gemsbok, steenbok and duiker rest partly on these identifications.

### 6. Other technologies, skills or innovations

- Global Positioning System (GPS) readings for every sampling point and every 5 km of the ground counts.

- IDRISI Geographic Information System (GIS) to measure each point’s distance from the nearest livestock waterpoint and to overlay impact zones on animal density maps.

- Aerial surveys from a Cessna 206, with counting strips about 200 m wide marked by rods and streamers on the wing struts.

- Trend-surface analysis and kriging to bring out a gradual pattern that could not be seen on the raw density maps.

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

- Hunting and poaching, rather than livestock grazing, appear to be the main reason gemsbok are scarce far beyond the livestock areas.

- Special game licence quotas are now supposed to be included in each hunting area’s total gemsbok quota.

- Springbok, concentrated in the same patches all year, could easily be depleted by hunting.

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

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

- Simple spoor counts, scored on a 0 to 5 scale, picked up the same gemsbok gradient as six aerial surveys and more advanced spatial statistics, and they were more useful than aerial surveys for small antelope such as steenbok and duiker, which are easily missed from the air.

- Building a local reference collection by following herds and identifying fresh tracks and dung is a practical way to anchor identifications, but the skill of the people identifying the spoor was not reported or tested.

## Cite this summary

Lawrence, K. 2026. Summary of: How are people affecting the distribution of less migratory wildlife in the southern Kalahari of Botswana? A spatial analysis (Verlinden and Perkins et al., 1998). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/1998-verlinden-perkins-how-people-affecting-distribution/