Vultures are highly mobile scavengers that cover vast distances on their daily foraging trips. Vultures display specific spatial and behavioral patterns when feeding on carcasses, and these behaviors are potentially identifiable in the tracks of geographical positioning system (GPS)- tagged birds. Thus, by applying algorithms to classify their movements it may be possible to create a carcass detection system using these spatial signatures. In this study I applied this approach to explore its efficiency in detecting carcasses and describing the diet of 12 GPStagged vultures in Botswana. I tracked five resident vulture species in Botswana using GPStrackers, these were: Cape Vulture (Gyps coprotheres n=2), Hooded Vulture (Necrosyrtes monachus n=2), Lappet faced Vulture (Torgos tracheliotos n=2), African White-backed Vulture (Gyps africanus n=4) and the White-headed Vulture (Trigonoceps occipitalis n=2). I attempted to search as many as possible of the feeding sites identified by the algorithms between January and June 2022. These potential feeding events were ground-truthed to confirm carcass presence. At each investigated site, any carcass found was identified down to species level. The ground-truthed sites constituted 22% (n=109) of all potential feeding sites identified within the study time frame (n=494). Carcasses were recorded at 62% (n=67) of these investigated sites. Carcasses associated with feeding events from all vulture species were located, except for Hooded Vultures. For this species only one site was checked, with no carcass found. At vulture feeding sites, carcasses from a total of ten taxa were identified. These were mainly wild ungulates (84%) with livestock making up a far lower proportion (16%) comprising of cattle (Bos taurus) and donkey (Equus asinus). Elephant (Loxodanta africana) was the most frequently identified taxon in the carcasses (31%; n=21) and was found at feeding sites of all four vulture species. Although sample sizes were relatively small for each species, it appeared that livestock was recorded only at Lappet-faced and White-backed Vulture investigated sites. Furthermore, Cape and White-headed Vultures had wildlife mostly giraffe and elephant at their investigated sites. Potential feeding events were ground-truthed on average 8 ± 8.35 days after the initial identification of the feeding event. The likelihood of detecting a carcass at these sites declined significantly with the amount of time lapsed from initial identification. The intercept of this relationship suggested that at the time of initial identification there was an 85% chance of detecting a carcass. Vegetation biomass at the sites, and the vulture species involved, had no influence on the likelihood of detecting a carcass. My study, on the application of GPS-tracking to understand the diet of a sample of African vultures in Botswana suggests that this technique is a useful approach but requires substantial resources in terms of personnel to investigate the identified feeding events as soon as these events are triggered, to derive a sufficient sample size to characterize the diets of these different vulture species.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:uct/oai:localhost:11427/38084 |
| Date | 12 July 2023 |
| Creators | Mpofu, Charles |
| Contributors | Amar, Arjun, Tate, Gareth and Maude, Glyn |
| Publisher | Faculty of Science, Department of Biological Sciences |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Master Thesis, Masters, MSc |
| Format | application/pdf |
Page generated in 0.0021 seconds