Small populations of animals are vulnerable to the consequences of breeding within a closed group – inbreeding depression and genetic drift lead to reductions in genetic variability, which in turn can give rise to the amplification of deleterious traits. Traditionally, managers attempt to minimise these effects by controlling the genetic structure via a manipulation of the paternal line, in the case of rhinos usually by translocation of breeding and sub-adult bulls. This strategy depends on having access to detailed pedigree data, and, in particular, some knowledge of the parentage of the offspring within the population. This information is particularly difficult to obtain in long-lived, free-ranging populations (particularly for rare and endangered species) where the identity of the father must be inferred from behavioural observations, and often the identity of the mother cannot be determined once juveniles have dispersed. In this thesis I present the results of a study to determine the parentage within a free-ranging, enclosed population of southern white rhinoceros (Ceratotherium simum simum) located on Ongava Game Reserve (OGR) in the north of Namibia. I used genetic techniques to obtain a genotype for each animal in the population, and then used detailed reserve management records from the period 1993-2009 to constrain the statistical process of parentage assignment. Using these different methods, I was able to assign both parents for 22 of 23 offspring with 80% confidence (16 of 23 at 95% confidence, mother only in 1 of 23), making this study the first to successfully complete a comprehensive parentage analysis in a free-ranging population of southern white rhinoceros. The key to the success of this study was a combination of accurate pedigree data and a complete set of genotype data. The parentage assignments allowed me to construct a complete lineage diagram for each of the founder matrilines, and further analyse the status and reproductive success of the population. OGR’s southern white rhinoceros population is expanding at close to 14%, well over the expected maximum growth rate for rhinoceros metapopulations (9%). The mean inter-calf interval is about 2.2 years and average age at first parturition is 6 years, indicating good fecundity. Conception is strongly seasonal, occurring mainly (89%) during the rainy season. The current management practice is to replace dominant bulls after their breeding tenure, and also to remove all young bulls from the population before they reach breeding age (with the aim to restrict potential inbreeding). Only one calf of seven in the F2 generation is inbred. My analysis indicates that, at least in the 2006 cohort of eight calves, founder females bred only with the founder male, while offspring females (F1 generation) bred only with introduced males. This suggests some form of mate selection leading to inbreeding avoidance within the population. Finally, there is some evidence that certain matrilines exhibit/experience different reproductive potential (daughters in one matriline exhibit longer inter-calving intervals and male bias in calf birth sex ratios). These details indicate that genetic data provide valuable information for management. When reviewing management decisions to date, I found that one of the founder bulls was the more successful in terms of calves sired (10 of 13). Management had, however, selected the other founder bull for removal by sale based on the assumption that he was behaviourally dominant and territorial and therefore likely to have been more successful at breeding. I also found that introduced bulls were breeding successfully before they appeared to have established territories. Thus these findings challenge the assumption that male white rhinoceros reproductive success is related to dominant, territorial behaviour. I conclude that in order to optimally manage small, free-ranging enclosed populations of southern white rhinoceros it is essential to have reliable and accurate pedigree data (this includes a methodology for identifying individual rhino), as well as genetic data for the entire population. I recommend that conservation management programmes for rhinoceros populations incorporate both genetic and demographic data. This will allow for the development of white rhinoceros population management strategies that attempt to optimize genetic diversity and population health, and benefit the establishment of new, robust populations. Translocations of animals are an important aspect of meta-population management of rhinoceros and data which provides accurate insight into the true mating system and reproductive success within a population allows for the correct selection of individuals for this process. Copyright / Dissertation (MSc)--University of Pretoria, 2012. / Zoology and Entomology / Unrestricted

Philosophiae Doctor - PhD / The survival of free-ranging (in situ) African elephant and Southern white rhinoceros populations are currently being challenged on a daily basis in Africa. Reproductive health is considered a vital component of species conservation. Conservation of the last mega land mammals may ultimately require intervention by breeding management or combined with assisted reproductive technologies (ART). There is a strong case for gathering baseline information, both physiological and biological, of any species, as opportunities arise. During this study a total number of 21 ejaculates collected over two seasons from 12 free-ranging African elephant bulls were characterised, as well as 10 ejaculates collected from 10 free-ranging Southern white rhinoceros bulls from two populations. Ejaculates were collected from adult bulls by means of electroejaculation under anaesthesia. Routine semen analysis was combined with Computer-aided sperm analysis (CASA), Computer-aided sperm morphology analysis (CASMA), Transmission electron microscopy (TEM) and Genomics as diagnostic tools. Additionally, sperm functionality within different media was investigated and sperm subpopulation classification according to the motion pattern displayed. The results presented is based on the evaluation and classification of ≈ 45 000 individual African elephant spermatozoa and ≈ 18 000 individual Southern white rhinoceros spermatozoa. The average elephant ejaculate contained a total number of 47 x 10⁹ spermatozoa (volume of 56 ± 38mL x concentration of 818 ± 750 x 10⁶/mL) that recorded a total motility of 81 ± 29% of which 62 ± 26% were progressively motile. CASA recorded velocities for curvilinear velocity (VCL 241 ± 58μm/s), straight-line velocity (VSL 173 ± 181μm/s) and average path velocity (VAP 201 ± 54μm/s), and kinematics at straightness of track (STR 86 ± 85%), linearity of track (LIN 67 ± 16%), amplitude of lateral head displacement (ALH 4 ± 0.75μm) and beat cross frequency (BCF 21 ± 3Hz). Structural analysis revealed 68 ± 11% of the spermatozoa were viable (intact plasma membrane) and 77 ± 11% maintained acrosome integrity. Ejaculates contained 55 ± 14% morphologically normal spermatozoa, CASMA measured sperm head lengths at 6.83 ± 0.26μm and width 3.32 ± 0.18μm (total head area of 20.17 ± 1.96μm²) of which 38.95 ± 0.92% is covered by an acrosomal cap. The average rhinoceros ejaculate contained a total number of 1.1 x 10⁹ spermatozoa (volume of 24 ± 24mL x concentration of 83 ± 96 x 10⁶/mL) that recorded a total motility at 82 ± 8% of which 28 ± 23% were progressively motile. CASA recorded velocities for VCL (85 ± 29μm/s), VSL (44 ± 25μm/s) and VAP (69 ± 30μm/s, and kinematics at STR (63 ± 14%), LIN (51 ± 16%), ALH (2 ± 0.16μm) and BCF (16 ± 6Hz). Structural analysis revealed 73 ± 10% of the spermatozoa were viable (intact plasma membrane) and 76 ± 4% maintained acrosome integrity. Ejaculates contained 62 ± 14% morphologically normal spermatozoa, CASMA measured sperm head lengths at 5.5 ± 0.17μm and width 2.9 ± 0.19μm (total head area of 14.8 ± 1.43μm²) of which 36.3 ± 0.59% is covered by an acrosomal cap. Based on a Boolean argument and CASA data exploration it was possible to derive elephant and rhinoceros CASA cut-off criteria to sort between activated and hyperactivated motile spermatozoa. For the genomic component of this study, the CatSper1 (Loxodonta africana) gene was identified,sequenced and verified in a free-ranging (natural) African elephant population. Multivariate analysis(MVA) was applied to examine the associations between the semen and sperm parameters and the traits they accounted for in this study. Our understanding of wildlife reproductive sciences can substantially progress as the analytical techniques applied and the combination thereof is expanded. This investigation presents a new set of comprehensive semen and sperm threshold values for future investigations.