Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Afrique du Sud → 2009 → Habitat suitability assessments for sable antelope

University of the Witwatersrand (2009)

Habitat suitability assessments for sable antelope

Chirima, Johannes George

Titre : Habitat suitability assessments for sable antelope

Auteur : Chirima, Johannes George

Université de soutenance : University of the Witwatersrand

Grade : Doctor of Philosophy 2009

Relationships between occurrence of a species and features of habitats occupied are central to establish factors that influence its distribution. Within large protected areas extinction processes may cause retractions of species distributions to areas that are still suitable or to locations least affected by a negative influence. The aim of the project was to identify factors that influence the suitability of areas where sable antelope occur. Climate and geographic barriers have overriding influences over biotic factors to identify regions that lie outside a species range. Abiotic factors (e.g. geology and rainfall) indicate places with environmental conditions that allow a species to persist (spatial extent of a fundamental niche). However, biotic interactions can constrain occupation to a limited proportion of those conditions (subset of fundamental niche). I used aerial census data (1977-93) in Kruger National Park to : (1) model distribution patterns commonly exhibited by large ungulate species with the objective of identifying methods most suitable for assessing different aspects of species distributions ; (2) assess how distribution patterns of 12 antelope species have apparently changed since around 1960 and how these changes may be related to sable distribution shifts or abundance decline ; (3) assess whether a climate effect could have caused contractions of range and abundance of sable and other rare antelopes and (4) identify features that restricted a widespread distribution of sable in KNP using logistic regression models. In chapter 2, I compared and contrasted performance of LoCoH and kernel methods for constructing distributions for species exhibiting (i) wide and continuous distributions with a few gaps, (ii) broad distribution with local concentrations and absences, (iii) linear distribution pattern associated with rivers, and (iv) a patchy distribution pattern. The methods have valuable capabilities for assessing different objectives of species distributions. The type of spatial distribution exhibited by a species influences the performance of these methods. This contrasts generalizations from home range studies that suggest superiority of one method over the other. The LoCoH method tends not to include areas where a species was not recorded. In contrast, kernel method exhibited the opposite bias. However, their differences were not large enough to lead to a diverse interpretation of range extents or occupancy patterns. Automatic procedures of choosing h appeared not adequate for mapping distribution patterns of species that occur in patches where outlines of outer boundaries are not clearly defined or for those species exhibiting clumped occurrences in places and widespread occurrences elsewhere. A different h value may be necessary for each section of such a distribution with fixed kernel method. This is achieved by dividing a study area into separate sections and mapping the ranges independently. The LoCoH is suitable for indicating gaps and/or fine-scale range shifts. However, LoCoH method may have to be applied with caution for species exhibiting continuous distributions because there is a possibility of emphasizing unimportant gaps. Despite the fact that distribution patterns around 1960 were vaguely complied, it appeared that common species have increased occupation of northern half of KNP and several species (impala, buffalo, wildebeest, warthog, and waterbuck) have been sighted during dry season (1980-1993) in areas indicated around 1960 as wet season range. The 1980-1993 distribution of impala, warthog, and waterbuck appeared more widespread away from rivers than around 1960. Distributions of sable, tsessebe, eland, and roan contracted in northern half and in central region of KNP. Fences that blocked migrations of wildebeest and zebra outside the park to the west of central KNP appeared associated with distributional changes of herbivores in this area. Augmenting surface waterpoints was a key influence in expansions of distributions of common species into northern half of KNP and for occurrences of some species during dry season in areas previously used during wet season. The contractions of distributions of rarer antelopes occurred concurrently with expansions of common species into northern half of KNP where rare antelopes mainly occur. The above suggests that some areas of northern half of KNP may have become less suitable to rare antelopes. Despite that the exact influence of climate on rarer antelopes could not be established, distribution pattern changes were characteristic of an influence consistent with that of climate. Range contractions were evident for all three species (sable, roan and Tsessebe), associated with local herd extirpations, especially following the severe 1991/2 drought. Herds of sable, roan, or tsessebe that occurred in isolated locations disappeared and ranges contracted even in the relatively wetter southern section of the park. Sable herds persisted in discrete patches after a widespread contraction of their formerly contiguous range in northern section. Sable prevalence was highest on nutrient poor granite and sandstone rather than nutrient rich basalt and gabbro. Distances from perennial water sources did not have overriding influences on where sable herds occurred. Sable prevalence was higher in mopane savanna woodland and sour bushveld than shrubland, dense bush savanna, or grassland with few trees. Sable herds were prevalent in localities that exhibited relatively low predation risk and low effects of competition from abundant grazers, implying that at the edge of a range, interactions involving biotic factors appear more important than searching for areas which potentially have more forage resources. Predation risk appeared more influential to sable distribution than competition. Findings showed that biotic factors strongly modify effects of abiotic factors on where rare and sedentary species establish.


Version intégrale (4,33 MB)

Page publiée le 6 avril 2011, mise à jour le 23 juin 2017