Includes bibliographical references. / The development of phytoplankton assemblages in relation to environmental variables was investigated in Lake Chivero between 2003 and 2006. The study showed that there has been a decline in dominance of cyanobacteria in Lake Chivero, instead phytoplankton development exhibited two states: (i) a clear state with lesser dominance of Microcystis and proliferation of cryptophytes and chlorophytes (eukaryotic algae) and (ii) a turbid state with domination by Microcystis and gradual exclusion of other species.

Includes bibliographies. / A study was made of two Redbilled (Green) Woodhoopoe Phoeniculus purpureus populations spanning an eight year period (1981-1988) and 258 'flock-years', in the eastern Cape Province of South Mrica. The main objectives of the study were to investigate (1) why woodhoopoes live in groups; (2) why nonbreeders do not breed; and, (3) why nonbreeders provision young that are not their own? Ecological and demographic data were gathered in addition to detailed behavioural observations of 54 woodhoopoe flocks. The following experimental manipulations were performed: (1) breeders were removed from flocks to (a) monitor dispersal patterns and restructuring of flocks; and, (b) observe behavioural reactions by remaining birds; (2) cavity availability was (a) decreased, to enable quantification of availability; and, (b) increased, by addition of nest/roost boxes to an area which supported no permanent woodhoopoe territories; and, (3) stimuli, associated with the food provisioning response of adult birds, were manipulated to investigate the evolutionary basis of allofeeding behaviour Variability in social and reproductive behaviour reflects environmental selection pressures, in the form of roost-cavity availability, with a reduction in cavity availability leading to increased group size. The group-territorial social system and high level of inbreeding of Redbilled Woodhoopoes have evolved primarily in response to environmental constraints on dispersal, rather than by particular benefits that arise from group living. Therefore, the habitat-saturation hypothesis best explains group living of woodhoopoes. Behavioural dominance hierarchies ensure that dominance relationships are well-defined among potential competitors (for breeding status), and thereby minimize disruption to flock cohesion upon the death (or removal) of a breeder. If competition for a breeding vacancy arose at the time of the breeder's death, the resultant delay in occupancy of the breeding vacancy would increase the likelihood of competition from unrelated birds. The establishment of such hierarchies is therefore adaptive in the context of the direct component of kin selection. The presence of nonbreeding helpers do not increase fledgling success, breeding frequency, survivorship (of any age, sex or social class), or number of breeder-offspring produced. Because no unambiguous indirect fitness benefits could be shown to result from helping behaviour (specifically allofeeding), I propose that the unselected (misdirected parental care) hypothesis is a viable alternative to the 'functional hypotheses.' This hypothesis is supported by observations/manipulations of misdirected food provisioning by both breeders and helpers.

The aim of this research is to describe the migration and moult strategies of the Curlew Sandpiper in relation to the different migratory routes and non-breeding areas of the species, focusing on the carry-over effects which conditions at the breeding grounds have on the timing of post-breeding migration and, subsequently, on the timing of moult at the non-breeding areas. Part of the study also describes the adaptive feature of moult in relation to migratory distance and environmental conditions at the non-breeding areas.

Includes bibliographical references. / This thesis deals with the development and application of a new tool to assess the conservation significance of wetland avifauna. Termed the Waterbird Conservation Value (WCV) score, this method sums the proportion of each species' count relative to the global 1% threshold level for each species; the value obtained measures the overall conservation importance of the wetland to waterbirds.

Includes bibliographical references. / Coral reefs are globally in a vulnerable state, because of both human impacts and environmental alterations. An understanding of coral reef ecosystems and the ability to detect changes in the reef environment early on are necessary to apply effective conservation. The co-evolution of coral and coral feeding fishes means that some species may exhibit interactions useful for indicating reef health, acting as an early warning system. Chaetodontidae (Perciformes) are one such family, containing coral dependent (obligate corallivore) species. This concept is known as the Butterflyfish Indicator Hypothesis (BIH). Through the collection of baseline data for butterflyfish abundance, diversity and species composition, the possible applicability of the BIH was investigated within Marine Protected Areas (MPA¡¯s) of iSimangaliso Wetland Park, on the north-eastern coast of South Africa. Other factors possibly impacting butterflyfish distribution were also investigated in this region. These factors were level of MPA protection (sanctuary or partially protected), reef rugosity and site effects. 78 transects across 13 dive sites were swum using Underwater Video Census (UVC) for data capture. Reef rugosity displayed no impact on any of the sampled factors, whilst site and protection level did. Protection was the factor with the greatest influence on butterflyfish species richness (df = 77, t = 2.85, p < 0.005), evenness of spread (df = 77, t = 1.79, p < 0.05) and total abundance, as revealed by the General Linear Model (GLM) with the lowest Aikake¡¯s Information Criterion (AIC) (y = ¥â&#8338; + ¥â©û (protection)). None of the factors altered Chaetodon meyeri distribution significantly, the only observed obligate corallivore, for which low abundances were observed. Differences in total butterflyfish abundance and species diversity appear to be particularly useful in revealing human impacts on coral health. The lack of C. meyeri, likely accounted to low abundances of its preferred coral Acropora (Scleractinia: Acroporidae), suggest that the BIH is not useful in this region.

Bibliography: pages 163-179. / Spirostreptida millipedes comprise three families, the Harpagophoridae, Spirostreptidae and Odontopygidae. They are polygynandrous. Males transfer sperm via species-specific accessory genitalia called gonopods, that comprise three components, two of which, the emote and telopodite, are involved in processes of sperm transfer. The emotes function to translocate ejaculates from the penes to the vulvae, where they are stored. A delay between insemination and fertilisation provides an arena for syn- and postcopulatory sexual competition. These operate at the gametic level via sperm competition and cryptic female choice. By combining studies of genital form and function with single and double mating experiments, this study elucidates these processes in some southern African Spirostreptida millipedes.

Bibliography: pages 316-331. / This thesis attempts to give an understanding of intertidal conditions in the Cape Peninsula in terms of spatial and seasonal variations in rocky shore communities and falls into two distinct parts: 1. A general survey of intertidal conditions and an investigation of the interactions of certain physical factors on rocky shore communities. 2. An investigation of seasonal changes in the populations of certain abundant algae and gastropods at Dalebrook.

Bibliography: pages 132-140. / The present study sought to make use of conditions during the formation of Lake Kariba, which resulted from the impoundment of the Zambezi. Three important factors led to changes in the ecology of the Kariba Basin, an area of 2,109 sq. miles, inhabited by primitive tribesmen and supporting extensive populations of large game mammals. In chronological order these were: i) the evacuation of the native inhabitants; ii) the preparation of fishing grounds; and iii) the inundation of almost the whole floor of the Central Zambezi Valley. The removal of the people resulted in large areas, especially those on or adjacent to fertile alluvium, which had been over-populated, being fallow until either cleared for fishing or submerged by the lake. The bush clearing of selected areas designated for fishing-grounds resulted in the destruction of most vegetation, while the filling of the Lake transformed a terrestrial environment, bisected by a seasonally fluctuating river, into a large static body of water within four and a half years. The habitat thus eliminated was typical of dry marginal areas in Rhodesia, the remaining strongholds of large mammals in the territory. This paper aims at describing some of the effects of the formation of lake Kariba on large mammals and their response not only to flooding, but also to the over-populated conditions which arose on islands on which they became marooned. In so doing it is, however, necessary to touch on the influence of past human settlement on the distribution and abundance of large mammals and on the effects of the clearing of incipient fishing-grounds. It is also convenient to mention briefly some of the most obvious responses of other groups of animals, as they showed certain similarities with those of mammals.

This study addresses the extrinsic factors which have shaped the evolution and maintenance of sociality in the African mole-rats. Specifically, the common mole-rat was used as a model to assess the Aridity Food-Distribution Hypothesis (AFDH), as an explanation for the evolution of bathyergid sociality. The AFDH correlates mole-rat sociality with habitat aridity and the pattern of food distribution. Aspects relating to ecological constraints, foraging behaviour, population demography, reproductive biology and aggressive behaviour were compared between an arid and a mesic population of C. h. hottentotus, to assess how inter-habitat divergence in ecological attributes has influenced social behaviour within these two populations. In evaluating the AFDH as an explanation for the evolution of sociality within C. h. · hottentotus two broad questions were addressed: (1) do the assumptions of the AFDH hold true i.e. do arid and mesic habitats exhibit ecological differences, specifically with regard to the pattern of resource dispersion and the energetic costs of foraging, which influence foraging risks and consequently the costs of dispersal? and (2) do these inter-habitat differences have implications for bathyergid social evolution i.e. do the common mole-rat populations inhabiting arid and mesic areas exhibit regional differentiation in social behaviour? Substantial inter-site divergence in ecological characteristics, notably climate and resource attributes, were revealed in this study. Rainfall at the arid site was markedly lower and more sporadic, and evaporation levels significantly higher, than at the mesic site. Moreover, thermal constraints were more limiting at the arid site. These features will greatly elevate the costs of soil excavation and the risks of hyperthermia, severely restricting the occurrence of suitable burrowing opportunities at the arid locality. Consequently, foraging will be severely constrained in this area. At the mesic site, higher, more predictable rainfall, low evaporation rates and reduced thermal constraints will translate into more suitable burrowing opportunities for much, if not all, of the year. Regional differentiation in food resource characteristics was also evident. Although geophytes were clumped at both study localities, the density of geophytes was lower and the distance between geophytes or geophyte clumps concomitantly greater at the arid relative to the mesic site. Differences in resource dispersion in turn influenced the patterns of foraging. In response to the low geophyte density and associated longer foraging distances, burrow systems were notably longer and more linear at the arid site. Furthermore, food storage and in situ harvesting were essential components of cooperative foraging in C. h. hottentotus as they minimised the risks of starvation, particularly in arid habitats. Thus, resource characteristics together with the climatic restrictions on burrowing in arid areas may have a marked impact on foraging behaviour, imposing severe constraints on the mole-rats occurring there and ultimately shaping their foraging responses. Together, these factors satisfactorily account for the underlying premise of the AFDH, that arid and mesic habitats exhibit ecological differences with regard to the pattern of resource dispersion and the energetic costs of foraging, which are likely to influence foraging risks and the costs of dispersal. In evaluating the AFDH, the second question which needed to be addressed was whether the study populations exhibited divergence in their social behaviour. The populations revealed no differences in absolute group size or in reproductive characteristics which were related to the effects of aridity per se. However, distinct inter-population divergence was readily apparent in phenotypically plastic traits such as dispersal behaviour and xenophobia. Clear differences were evident between the arid and mesic sites in both the quantitative and qualitative nature of dispersal; dispersal was markedly constrained at the arid site and colonies demonstrated greater temporal stability, with more predictable temporal group membership. The ecological constraints on successful foraging at the arid site will curb opportunities for dispersal and promote cooperation in the C. h. hottentotus occurring there. Colony members should therefore maximise their inclusive fitness by natal philopatry, delayed dispersal and cooperative foraging. Inter-site differences were also apparent in the response of colony members to foreign conspecifics. Common mole-rats from the arid site were markedly more xenophobic than those from the mesic site, and aggressively rejected foreigners. For arid-occurring populations, the fitness penalties for failing to exclude foreigners from the colony burrow system and associated resources, will be more severe than for mesic-occurring populations, resulting in heightened levels of xenophobia. Again, colony cohesion and cooperation in arid areas are essential to individual survival and inclusive fitness. The regional differences in dispersal patterns and xenophobia revealed in this investigation may reflect adaptive variation in social behaviour between the study populations, and the results suggest that delayed dispersal and cooperation may be more crucial to individual survival in arid than in mesic areas. As such these findings provide support for the underlying contention of the AFDH that ecological constraints on foraging in arid areas have promoted a greater degree of social elaboration in mole-rats occurring there. This study provides persuasive support for the AFDH as an explanation for the adaptive significance of social behaviour and cooperation in the common mole-rat, and together with other investigations, suggests that the AFDH provides a valid explanation for the evolution of group-living in the Bathyergidae.

Bibliography: p. 109-122. / This thesis explores the relationships between commercial data and those obtained from research surveys in an attempt to broaden the database available for management, and to cover seasonal and inter-annual changes in density estimates of Cape hake (Merluccius capensis). It also attempts to explain hake variability of in terms of environmental indices based on satellite remote sensing, carefully chosen to reflect underlying oceanographic processes. The influence of various factors on catch rates (CPUE) is investigated using a general linear model.