# Indexing small mammalian carnivores in the southern Kalahari, South Africa

> In the Kalahari, the wind wipes the dunes clean every day, leaving a fresh page for each night’s tracks. With a Khoisan tracker reading those tracks over four years, researchers monitored 10 small carnivore species across farms with increasing shrub encroachment, and found that counting each set of tracks, not just noting presence, revealed population changes, and even breeding success, that a simpler index missed.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/2008-blaum-engeman-indexing-small-mammalian-carnivores/

## Citation and access

- **Authors:** Niels Blaum, Richard M. Engeman, Bernd Wasiolka and Eva Rossmanith
- **Year published:** 2008
- **Journal or source:** Wildlife Research
- **Full citation:** Blaum, N., Engeman, R. M., Wasiolka, B. and Rossmanith, E. 2008. Indexing small mammalian carnivores in the southern Kalahari, South Africa. Wildlife Research 35(1): 72–79.
- **DOI:** 10.1071/WR07023
- **Access status:** Hosted (freely available elsewhere)
- **PDF:** https://www.originalwisdom.com/wp-content/uploads/2026/11/2008_Blaum_and_Engeman_etal_indexing-small-mammalian-carnivores.pdf

## Tags

- **Species:** Mammals > African wild cat; Mammals > Bat-eared fox; Mammals > Black-backed jackal; Mammals > Black-footed cat; Mammals > Cape fox; Mammals > Slender mongoose; Mammals > Small-spotted genet; Mammals > Striped polecat; Mammals > Suricate; Mammals > Yellow mongoose
- **Region:** Africa > South Africa
- **Tracking:** Track counts; Tracking; Tracks; Tracks and signs identification
- **Behaviour and ecology:** Behaviour
- **Methods and technology:** Indigenous knowledge; Linear regression; Population trends
- **Data type:** New field data
- **Observer qualification:** Reported but not validated
- **Discipline:** Behaviour and ecology; Indigenous knowledge; Wildlife tracking
- **Year published:** 2008

## Detailed keywords

- **Species:** Studied (10 small carnivores, names as given in the paper): yellow mongoose (Cynictis penicillata), African wild cat (Felis lybica), striped polecat (Ictonyx striatus), bat-eared fox (printed as “Octocyon megalotis”), Cape fox (Vulpes chama), small-spotted genet (Genetta genetta), black-backed jackal (Canis mesomelas), small-spotted cat (Felis nigripes, also known as the black-footed cat), slender mongoose (Galerella sanguinea) and suricate or meerkat (printed as “Suricata suricata” in Table 1). In the study area but not studied: cattle and sheep (livestock on the farms). Vegetation named: camel thorn (Acacia erioloba), grey camel thorn (A. haematoxylon), shepherd’s tree (Boscia albitrunca), Rhigozum trichotomum, black thorn (A. mellifera), and grasses (Stipagrostis, Eragrostis, Schmidtia kalahariensis); dune crests mainly Stipagrostis amabilis. The paper gives scientific names only for the plants; the common plant names here are ours and can be removed. Inconsistencies in the paper: Tables 2 and 3 label African wild cat “AW” while Table 1 uses “AWC”; Table 2 gives suricate on Farm 1 in 2002 as 0.15 ± 0.16, where the standard error is larger than the value and the value does not fit the 1/32 steps of the other entries.
- **Location and habitat:** Five cattle and sheep farms in the southern Kalahari rangeland south of the Kgalagadi Transfrontier Park, between Twee Rivieren and Askham (26°15′S, 20°35′E), Northern Cape Province, South Africa. Western Kalahari Thornveld savanna: scattered trees and shrubs in a grassy matrix. Rainfall about 174 mm a year, mostly January to April; temperatures from −10.3°C to 45.4°C. Shrub cover on the farms: 4%, 9%, 15%, 18% and 26% (stated as 25.5% for Farm 5 later in the Methods). Farms 7.5 to 40 km apart; stocking rates from 2 to 6 livestock units per 100 ha. Transects on sand along dune crests with under 5% plant cover.
- **Topics:** passive tracking index; track intrusion counts versus presence/absence (binary) data; relative abundance indices; low-cost monitoring of nocturnal and secretive carnivores; indigenous Khoisan tracking knowledge; shrub encroachment, overgrazing and habitat degradation; habitat fragmentation; population trends; patchy distribution; reproductive success shown by tracks of adults with young; labour costs of monitoring; region-wide monitoring on private farmland.
- **Statistics or models:** Binary index (BI: mean proportion of transects with tracks per subplot) and track intrusion index (TII: mean number of track intrusions per transect, after Engeman 2005), each with variance; linear random-effects model with variance components by restricted maximum likelihood (REML) in SAS PROC VARCOMP; correlation coefficients between the two indices; linear regression of index values against shrub cover (Fig. 2: African wild cat 2005, TII R² = 0.86, P = 0.015, BI R² = 0.65, P = 0.055; Cape fox 2001, TII R² = 0.85, P = 0.016, BI R² = 0.93, P = 0.005). Inconsistencies in the paper: the Discussion says the yellow mongoose on Farm 4 showed decreasing (BI) versus increasing (TII) trends “(Fig. 1)”, but the Results and Tables 2 and 3 show the reverse (BI steady or rising, TII falling), and Fig. 1 shows Farm 1 and does not include the yellow mongoose. The text says both indices detected the effect of shrub cover, but the BI trend for the African wild cat in Fig. 2 has P = 0.055. Table 2 is titled “transect intrusion index” while the text uses “track intrusion index”.
- **Data type:** New field data: track surveys in the rainy season (February to May) of 2001, 2002, 2003 and 2005 (four years in a five-year period); 640 transect surveys in total (4 years × 5 farms × 4 subplots × 8 transects).

## Synopsis

### 1. What they did

In the southern Kalahari, heavy grazing has led to shrub encroachment, which breaks up the savanna into fragments. Small carnivores are vulnerable in fragmented landscapes, but they are hard to count because they are few, mostly nocturnal and secretive. Wildlife managers in Africa often lack equipment, money and trained staff, so the authors wanted a simple, cheap method that could monitor many species at once. They combined a passive tracking survey with the knowledge of a Khoisan tracker.

- The farms. They chose five cattle and sheep farms with shrub cover from 4% to 26%, forming a gradient of habitat degradation.

- The transects. Each farm had four subplots, at least 2 km apart, and each subplot had eight sand transects (5 m × 250 m) along dune crests. Strong winds almost every day around midday wiped old tracks away, so the sand did not need to be prepared, and no bait or lures were used.

- Reading the tracks. Each transect was checked for 2 hours, starting 1 hour after sunrise, for small carnivore tracks made in the previous 18 hours. Andrew Kruiper, a Khoisan Bushman tracker from the Kgalagadi Transfrontier Park, identified the tracks and the team recorded how many times each species had crossed onto each transect (track intrusions).

- Two indices. From the same data they calculated a binary index (BI), based only on whether a species’ tracks were present on a transect, and a track intrusion index (TII), based on the number of intrusions. They compared the two and related them to shrub cover.

Surveys ran in the rainy season (February to May) in 2001, 2002, 2003 and 2005.

### 2. What they found

Species detected

- Tracks of 10 small carnivore species were recorded, and every species was found on every farm, although not in every year.

- The most common tracks were of yellow mongoose, African wild cat and striped polecat. The least common were small-spotted cat, slender mongoose and suricate.

Habitat degradation

- For almost every species and year, both indices were highest on Farm 1 (least shrub cover) and lowest on Farm 5 (most shrub cover), and the farms in between tended to follow the gradient.

- African wild cats and Cape foxes declined as shrub cover increased. For the less common Cape fox both indices showed a similar decline; for the more common African wild cat the TII showed the decline much more strongly.

- African wild cats showed a straight-line relationship with shrub cover, while small-spotted genets and bat-eared foxes appeared to peak at middle levels of shrub cover.

Comparing the two indices

- Over all farms and years, the two indices were strongly correlated for most species (0.55 to 0.96). The low value for slender mongoose arose partly because correlations could not be calculated on three farms with all-zero data.

- For rare species, the two indices were nearly the same, because a transect seldom had more than one intrusion.

- The TII usually had larger standard errors than the BI.

- For species that were common but patchy, the indices could point in opposite directions. On Farm 4, the BI for yellow mongoose stayed steady or rose (0.38, 0.38, 0.44, 0.44 from 2001 to 2005), while the TII fell (1.06, 0.81, 0.44, 0.44), giving a correlation of −0.94. On Farm 1 the TII showed strong swings in African wild cat and striped polecat numbers that the BI did not.

Reading the tracks of families

The team often saw tracks of one or two large animals beside tracks of one to three much smaller ones at the same spot on a transect. The authors explain the falling TII for yellow mongoose on Farm 4 as a drop in breeding success, with adult numbers changing only slightly.

### 3. What they concluded

- This simple, inexpensive passive tracking method, with tracks identified by a Khoisan tracker, was able to detect 10 small carnivore species and monitor changes in their relative abundance across a grazing gradient.

- Track indices cannot measure absolute density, but they can show relative abundance across farms.

- The TII is more sensitive to population differences and trends, especially for more numerous species that are spread unevenly. Because the surveys ran during the breeding season, when young travel with their parents, the TII may also track changes in breeding success and could hint at future population trends.

- The BI can save labour: once a track of the target species is found, the observer can stop checking that transect. It may suit projects with little money for labour, only one or two target species, or few expected tracks.

- Longer-term monitoring is needed to see whether lower breeding success leads to population decline or whether neighbouring animals move in.

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

Tracking was the core method. A Khoisan tracker identified small carnivore tracks (footprints) to species on wind-cleared sand transects and the team counted each species’ track intrusions. The paper also reports reading the tracks of adults and much smaller young together, which the authors use to interpret breeding success. No trailing is reported.

### 5. Tracker or observer certification

Reported but not validated. The tracks were identified by Andrew Kruiper, a Khoisan Bushman tracker from the Kgalagadi Transfrontier Park. The paper says that experienced Khoisan trackers readily identify animal tracks to species (citing other sources), but gives no certification or test of his skill. All results depend on his identifications.

### 6. Other technologies, skills or innovations

- Natural wind-cleared sand on dune crests served as tracking plots, with no surface preparation and no bait.

- A track intrusion index adapted from the general indexing approach of Engeman (2005), with variance components estimated by restricted maximum likelihood (REML) in SAS.

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

- Shrub encroachment caused by land use greatly reduces local carnivore numbers, while hunting pressure on small carnivores is low (except for black-backed jackals, which are killed because they prey on livestock).

- Monitoring carnivores across the region outside protected areas is a major challenge in the southern Kalahari, where well over 500 farms range from 3000 ha to more than 30 000 ha. This tracking method could make region-wide monitoring affordable.

- Because carnivores can serve as indicators of habitat degradation, monitoring them can show the effect of degradation on wildlife.

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

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

- The study shows how a single local tracker’s identifications can underpin a multi-year, multi-species monitoring programme. Because every number rests on those identifications, a test or certification of the tracker’s skill would strengthen this kind of study.

- Counting separate track intrusions, rather than only noting presence, kept information that changed the conclusions about trends. Reading adult and young tracks together added a further layer: a possible signal of breeding success.

## Cite this summary

Lawrence, K. 2026. Summary of: Indexing small mammalian carnivores in the southern Kalahari, South Africa (Blaum and Engeman et al., 2008). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/2008-blaum-engeman-indexing-small-mammalian-carnivores/