Department of Biology, Norwegian University of Science and Technology, 7491 Trondheim, NORWAY.
Large carnivores are particularly susceptible to negative anthropogenic effects. As human populations continue to encroach on wildlife habitats, protected areas may become increasingly important for large carnivore conservation. The African wild dog (Lycaon pictus) occurs at low densities and ranges over wide areas where anthropogenic activities account for a large proportion of mortalities. However, conserving wild dogs within protected areas presents several challenges that conservation managers need to address. In particular, wild dog populations occurring within the relative safety of protected areas are prone to sources of mortality associated with edge effects, intra- and interspecific competition, while fragmentation threatens population viability.
Land use along protected area boundaries is often characterised by an abrupt transition. Within the boundaries carnivores are protected and valued, but outside are regarded as a threat to humans their domestic animals, leading to their persecution. In addition, these transitional zones are associated with other threats such as accidental deaths inflicted by motor vehicles, contraction of disease from domestic dogs (Canis familiaris), and wire snares set by bush meat hunters. Edge effects are a major problem in the conservation of wild dogs, yet limiting their ranging behaviour poses a seemingly insurmountable challenge. However, wild dogs are territorial and the scent marks of neighbouring packs directly influence their ranging behaviour as neighbouring packs are avoided since encounters can be potentially fatal. We found that targeted deployments of translocated foreign scent marks influenced the ranging behaviour of an otherwise free-ranging wild dog pack. When exploring unfamiliar areas near or outside the protected area boundary, exposure to foreign scent marks resulted in the pack covering significantly greater distances and moving in the direction of their frequented range inside the protected area. Unlike wire fences, scent marks present a biologically relevant cue which, once detected, elicits a behavioural response. With further experimentation and a greater understanding of wild dog territoriality operates, this technique has the potential to reduce wild dog ranging into conflict-prone landscapes alongside wildlife reserves.
For the scent mark technique to be an effective management tool, wild dog packs should ideally be repelled by the presence of foreign scent marks. Scent marks are used to communicate territorial boundaries yet all studies of wild dog territoriality have to date reported that the degree of territorial overlap between neighbouring packs varies hugely. Attempting to maximize the effectiveness of managing wild dog ranging behaviour using foreign scent marks would necessitate an understanding of why, in certain instances, wild dog territorial boundaries are apparently ignored resulting in extensive transgression. We found that inter-pack relatedness was extremely important; unrelated packs had very low levels of overlap and these were restricted to the peripheral parts of territories. In contrast, related dyads overlap significantly more and even intensively utilized parts of the range, where the probability of encountering a neighbouring pack is far higher, were found to overlap. Pack size had little effect between related neighbours, but in unrelated dyads there was a tendency for larger packs to overlap more onto the territories of smaller packs. While Intraspecific competition can result in mortality, all available evidence suggests that aggression is greatly reduced between related packs. Establishing a territory alongside close relatives may therefore extend exclusive fitness benefits from the intra-pack to inter-pack level. The natural spacing system thus has the potential to affect the number of packs a given protected area can accommodate, as well as well as survival parameters through variations in the intensity of intraspecific competition. The findings are therefore important for the management of the species, specifically in intensively managed populations. For the purposes of using translocated scent marks to create a biological barrier, it is therefore important to use foreign scent marks from unrelated individuals as scent marks from related packs may not illicit the desired behavioural response.
The effects of interspecific competition on wild dog populations are, however, usually far greater than intraspecific competition. Lions (Panthera leo) and, to a lesser degree, spotted hyenas (Crocuta crocuta) can have major effects on wild dog populations and in combination account for more than 50% of wild dog deaths. As the dominant carnivore, lion densities are positively correlated with prey availability. Spatial variations in lion densities are therefore associated with specific landscape types and low lion density habitats are exploited by wild dogs, facilitating coexistence. Lions do not prefer hilly, mountainous and other forms of rugged terrain. With their decreased risk of lion encounter, we found than wild dogs actively selected rugged terrain when denning. Since wild dogs only reproduce once per year, the success of the annual event has a large influence on pack and population viability. Den site selection thus forms a primary defense strategy against predation.
Wild dogs are also vulnerable to interspecific competition outside of the denning period when ranging widely across their territory. Habitat selection should therefore attempt to minimize the probability of risky encounters with lions and hyenas. As case in point is the reported local extinction of a studied population of wild dogs in the Serengeti National Park (SNP), Tanzania. Both disease and interspecific competition were identified as contributing factors, yet the ultimate cause of the extinction has never been identified. Over a twenty-five year period the population declined before finally disappearing entirely. Although not well documented, the wild dog population did persist in the eastern parts of the ecosystem. Since the habitat characteristics are very different in these parts, we hypothesized that certain ecological features affected the distribution and availability of competition refuges. Given the dramatic increase in lion and spotted hyena populations during the wild dog population decline, competition refuges may have become increasing important to limit costly interspecific competition. A species distribution model was derived using occurrence records from the extant population within the ecosystem. We then used global positioning system (GPS) collar data from two wild dog packs recently reintroduced into SNP and overlaid the data onto the habitat suitability model. Habitat intensively utilized by wild dogs corresponded closely to habitat predicted as highly suitable, which was characteristically rugged. An analysis of movement data revealed that this habitat acted as competition refuges, the distribution of which was very limited within the SNP. However, habitat with an ability to provide wild dogs with respite from competition was abundant in the eastern parts of the ecosystem, where the population has managed to persist. The reported local extinction appears therefore to have instead been a range contraction; as lion and hyena populations increased, wild dogs were only able to survive in areas where they were not permanently exposed to high levels of competition. Topographic heterogeneity, and in particularly patches of rugged terrain, are therefore important should wild dogs successfully coexist in the presence of lion and hyena populations. While edge effects, intraspecific and interspecific competition are associated with protected areas and their immediate surroundings, many populations of wild dogs are located in geographically isolated populations. Population viability is greatly reduced in such instances. Wild dogs are, however, capable of dispersing long distances in the search of mates. In South Africa, the majority of wild dog populations occur in relatively small and geographically isolated populations. Most of these reserves have reached carrying capacity and a large proportion of the country’s total population occurs outside protected areas. We identified five priority conservation areas where habitat was relatively intact and where wild dogs had been sighted on numerous occasions, inferring some degree of connectivity to a source population. We assessed connectivity using different methodologies and using a dataset of occurrence records for both dispersing and resident packs of wild dogs. Despite the use of potentially subjective cost layer, current flow analysis provided the best connectivity analysis with greater potential to be incorporated into management plans. Our study shows that using habitat selection data from animals in resident populations, connectivity may be underestimated as dispersing individuals are often less habitat specific.