**4.3 Results**

When all the radio-telemetry locations (n = 847) were plotted on the map of the Mkhuze Game Reserve, they appeared scattered throughout the entire reserve, although the

Use of Telemetry Data to Investigate

*Broad Habitat Classification* 

*Refined Habitat Classification* 

*Acacia nilotica* low-closed

expected and available.

reintroduction program.

**4.4 Discussion** 

*System* 

*System* 

**Habitat Observations** 

Home Range and Habitat Selection in Mammalian Carnivores 297

*Terminalia sericea* savannah 155 0.187 0.048 0.093 + Woodland habitat types 279 0.336 0.058 0.379 0 Forest habitat types 269 0.324 0.056 0.291 0 Other habitats 10 0.012 0.013 0.014 0 Low-thicket habitat 49 0.059 0.029 0.096 - River floodplains & streams 68 0.082 0.034 0.127 -

*Terminalia sericea* savannah 155 0.187 0.039 0.093 +

woodland 134 0.161 0.037 0.049 + Acacia tall-open woodland 106 0.128 0.033 0.160 0 *Acacia tortillis* low-open woodland 49 0.059 0.024 0.064 0 Dry-closed woodland 122 0.147 0.035 0.167 0 Lebombo-wooded grassland 52 0.063 0.024 0.071 0 Open-woodland 72 0.087 0.028 0.084 0 Other habitats 10 0.014 0.013 0.014 0 Low-thicket habitat 49 0.059 0.024 0.096 - River floodplains & streams 68 0.082 0.027 0.127 - Riverine forest 13 0.016 0.013 0.076 -

Table 3. Comparisons of use and availability of various habitat types of an African wild dog pack reintroduced to Mkhuze Game Reserve based on telemetry data. Habitat types are shown in order of preference. Comparisons were made using the Neu Method and Bonferroni confidence intervals. Preference was determined with Bonferroni confidence intervals (*α* = 0.001) placed on use. "+" indicates used more than expected; "-" indicates used less than expected; "0" use according to availability or non-significant difference between

An important requisite for the success of any reintroduction program of mammalian carnivores is the elimination of the factors that initially caused the decline of the species. As previously mentioned the main factors explaining the decline of wild dogs in Africa include killing by humans, reduced prey availability, competition with other carnivores, loss and fragmentation of habitats, as well as infectious diseases. The wild-dog pack reintroduced into Mkhuze Game Reserve in 2005 was seen in this game reserve only until June 2008. Although it is suspected that some pack members moved out of the reserve onto neighbouring conservation parks or reserves, other individuals were found dead within or nearby Mkhuze Game Reserve. Possible causes for the deaths of these animals include snares, predation, hunting injuries and infectious diseases that are still occurring in the study area (personal communication, Mkhuze Game Reserve staff and veterinarians). Thus, the analyses discussed below contribute to a forensic-like study that we hope can provide some insights into the factors that may have played a role in the failure of this wild-dog

**Habitat Use proportion Expected** 

**Use proportion** 

**Preference Average SD** 

**(# locations)** 

number of radio-telemetry observations were higher near the southern border of the reserve (Fig. 4). A 100% MCP yielded a home range of 383.9 km2 while the 95% fixedkernel method resulted in a home-range of 377.8 km2. Thus, both 100% MCP and 95% fixed-kernel methods resulted in home-range sizes that were larger than the game reserve (~360 km2). In terms of centers of activity, the 50% fixed-kernel method resulted in a core-area of 103.6 km2, representing about 29% of the total area covered by the game reserve.

Fig. 4. Map of the Mkhuze Game Reserve showing 847 radio-tracking locations collected from a wild-dog pack. The solid line represents the 100% MCP home-range of the wild-dog pack calculated in this study.

Using the broad habitat classification system, wild dogs selected only for one habitat, *Terminalia sericea* savannah habitats while thicket habitat types and river floodplains were used less than expected based on their availability (Table 3). Using the most refined habitat classification system, our analyses indicated that out all woodland habitat types, *Acacia nilotica* low-closed woodland was the most important for wild dogs. Using this same classification system, it was also found that wild dogs clearly selected for *Terminalia sericea* savannah habitats (Table 3). These analyses also showed that the habitats that were used less than expected based on their availability were thicket habitat types, river floodplains and riverine forest. No significant preference was observed for any of the other habitat types.


Table 3. Comparisons of use and availability of various habitat types of an African wild dog pack reintroduced to Mkhuze Game Reserve based on telemetry data. Habitat types are shown in order of preference. Comparisons were made using the Neu Method and Bonferroni confidence intervals. Preference was determined with Bonferroni confidence intervals (*α* = 0.001) placed on use. "+" indicates used more than expected; "-" indicates used less than expected; "0" use according to availability or non-significant difference between expected and available.
