Bibliography: leaves 85- 97. / The spawmng habitats of anchovy (Engraulis capensis) and sardine (Sardinops sagax) in the southern Benguela upwelling ecosystem were characterized by comparing the data on abundance and distribution of eggs of the two species with environmental variables. Data were collected from two different survey programs: (a) 14 SARP (Sardine and Anchovy Recruitment Programme) cruises, conducted monthly for 2 spawning seasons between August 1993 and March 1994, and September 1994 and March 1995, off the west coast of southern Africa from Cape Columbine to Cape Agulhas; (b) annual November/December spawner biomass surveys conducted from 1984-1999 along the South African coast and covering the continental shelf between Hondeklip Bay on the West Coast and Port Alfred on the South Coast. A CalVET net was used to collect fish eggs at stations on a survey grid, and physical and biological data were collected concurrently with egg samples. Physical variables measured included sea surface temperature, nitrate concentration, water depth, salinity, current speed, wind speed and mixed layer depth, whereas biological variables included phytoplankton biomass (as 50m integrated chlorophyll a) and zooplankton biomass and production. The spawning habitats selected by anchovy and sardine were identified by constructing quotient curves derived from egg abundance data and individual environmental variables. Anchovy and sardine spawning probability with reference to the distribution of environmental variables was examined using overlay operation analysis. Relationships between eggs of the two species and the environmental variables were verified using multivariate co inertia analysis. Relationships among the environmental variables were examined through cluster analysis. During SARP surveys, single parameter quotient analysis indicated that the spawning habitats of these two species were most dissimilar in terms of water depth, sea surface temperature, current speed and zooplankton biomass, and most similar in terms of salinity and phytoplankton biomass. Coinertia analysis using all environmental variables as inputs showed a positive association between anchovy eggs and salinity and sea surface temperature, and a negative association between anchovy eggs and secondary production and phytoplankton biomass. Sardine eggs were strongly positively associated with phytoplankton biomass. During spawner biomass surveys, anchovy and sardine spawning habitats appeared to differ with respect to sea surface temperature, wind speed and current speed, and overlapped in terms of water depth, phytoplankton biomass, zooplankton biomass and production. Anchovy eggs were strongly positively associated with SST, salinity, mixed layer depth and zooplanktop production. Sardine eggs were strongly positively associated with current speed and zooplankton biomass.

Bibliography: leaves 78-86. / Research was carried out at the Walvis Bay/Swakopmund nursery during four visits between Aug/Sep and May/Jun 2000. The study aimed to quantify seasonal and tidal patterns of habitat use and responses to changes in prey abundance, and to investigate whether nurseries provide favourable foraging conditions and high energetic returns for juvenile oystercatchers.

Summary in English. / Bibliography: p. 221-242. / The cranes (Gruidae) and bustards (Otididae) are highly threatened taxa. Many species are on the brink of extinction due to anthropogenic factors. On the positive side, several species show a remarkable ability to adapt to man-modified environments. If the cranes and bustards are to survive through the coming century, research on the aspects of their biology directly relevant to understanding their conservation requirements is a high priority. Without a reliable and relevant base of knowledge, no successful conservation management plan can be devised for each threatened species. This thesis presents information on three threatened members of these taxa in southern Africa, the Blue Crane Anthropoides paradiseus, and Ludwig's Neolis ludWigii and Stanley's N. denhami stanleyi Bustards. This information furthers our understanding of their conservation status. The thesis begins with a detailed review of the phylogeny, general biology and conservation status of cranes. This is necessitated by the absence of any thorough and recent review being available. The review provides a point of reference for the research results on the Blue Crane presented in the following seven chapters. Research on the Blue Crane largely focuses on the southwestern Cape and Karoo regions. The past and present distribution of the species in southern Africa is outlined and related to various vegetation types and the effect of various anthropogenic influences. Its colonization of the southwestern Cape is a particularly interesting finding. The abundance of the species in various parts of South Africa is discussed, largely based on road counts and bird atlas data. The total population size in South Africa is estimated and the previously unrecognised importance to the species of the southern Cape coastal plain between Botrivier and Mossel Bay is highlighted. The scarcity of Blue Cranes in the western part of the southwestern Cape, compared with the southern part, is quantified. The reasons for this anomolous dichotomy in abundance betweep these two superficially similar areas are unclear and worthy of further investigation.

Bibliography: leaves 134-143. / This study was conducted with the overall objective of providing scientific infonnation relevant to the management of the gill and beach-seine net fisheries, particularly with respect to decisions about increased participation in the fisheries. Its primary aims were to provide quantitative estimates of total catch and effort, to assess the current and potential future importance of by-catch in the fisheries, to describe the socio-economic status of participants and to evaluate the management measures currently in effect.

Habitat fragmentation has different effects on species and communities, depending on a suite of life-history and population traits: some species are more vulnerable to the effects of fragmentation than others. Contrasting responses suggest there are particular species' attributes that make an organism more or less susceptible to the effects of fragmentation. Much research has focused on identifying which of these traits are the most useful indicators of a species' fragmentation-linked extinction risk. For example, body size, rarity, ecological specialization, matrix use, range size and turnover rate have all been linked with species extinction risk. Few studies have, however, attempted to explore the traits that predispose raptors to vulnerability from fragmentation. In this study, I compare the responses of two near-sympatric raptors (the Black Harrier Circus maurus and the Black Sparrowhawk Accipiter melanoleucus) to habitat fragmentation. On a broader scale, I use a simple model of susceptibility to fragmentation effects, and a sample of hawks (Accipiter spp) and harriers (Circus spp) in the family Accipitridae, to predict which species attributes are most likely to produce a negative response to habitat fragmentation. I then compare these predictions with the current global threat status of each species to test whether the model can predict threat status with acceptable accuracy.

Alien invasive fishes are a growing concern in inland water ecosystems around the world, as they are capable of causing serious damage, especially to indigenous fish populations. Mechanisms include direct predation on indigenous fish by alien predatory species, competition for food and space between native and introduced species, the introduction of alien parasites and pathogens, and general environmental degradation. The Cape Floristic Region (CFR) of South Africa, which is defined by a unique and highly diverse floral kingdom, is also home to a unique and highly threatened ichthyofauna. This ichthyofauna consists of relatively few species, but exhibits the highest proportion of endemicity in the country. Threats to fishes include habitat destruction through bulldozing and water extraction, water quality degradation, restriction of migration by dams and weirs, and alien invasive fishes. The North American smallmouth bass (Micropterus dolomieu Lacepede) has long been regarded by nature conservators as the most threatening invasive species, and much anecdotal and survey data suggests that this species has had a major role in the depletion and extirpation of indigenous fish populations throughout the CFR. However, no study has ever properly quantified this impact relative to other potential threats that could have precipitated the perceived indigenous fish losses. This is a problem, as it makes the implementation of active control measures difficult to justify to a sceptical public and potential sponsors. The Rondegat River is a tributary of the Olifants River, which rises in the Cedarberg Mountains and flows into the Clanwilliam Dam reservoir. It is partially invaded by M dolomieu, which have penetrated the lower quarter of the river up to a waterfall barrier. This river is home to five species of indigenous fish, including the Clanwilliam yellowfish (Labeobarbus capensis Smith), Clanwilliam redfin (Barbus calidus Barnard), fiery redfin (Pseudobarbus phlegethon Barnard), Clanwilliam rock catfish (Austroglanis gilli Barnard) and the Cape galaxiid (Galaxias zebratus Castelnau). This project was designed to quantify the impact of M dolomieu on these species relative to the alternate potential impacts of physical habitat degradation from agriculture and alien invasive riparian trees, and of food availability. Seasonal surveys were conducted at eight sites on the river in September, October and November 2003, and in April 2004. Four sets of riffles and pools were selected upstream of the waterfall barrier, and four below. Quantitative electrofishing was used to survey fish in riffles, while snorkelling surveys were conducted in pools. All fish species were also caught with seine and tyke nets for dietary analyses. Physical habitat variables were measured at each site, and used to assess changing habitat quality between the sites. Invertebrate samples were also taken along with visual abundance estimations of other food types, to gauge food availability. Fish surveys revealed the loss of B. calidus, P. phlegethon, A. gilli and G. zebratus at bass-invaded sites. Galaxias zebratus was only found in the upper reaches of the river, and so was possibly never common in the lower river. Labeobarbus capensis, while still below the waterfall, appeared to be suffering from near-total loss of post spawning recruits. Discriminant function analyses revealed sedimentation to be a key factor of habitat degradation that characterized invaded sites. However, linear regressions between habitat variables and indigenous fish densities indicated sedimentation to not be a significant negative influence on indigenous fish distributions. Although sedimentation did not appear to influence A. gilli densities at non-invaded sites, it is highly likely that it increased the vulnerability of this species to M dolomieu in the lower river, by removing benthic cover used to avoid predation. In the case of all species, food availability did not seem an important factor in dictating fish distributions. Consequently, predation by M dolomieu was confirmed as the critical mechanism behind the loss of B. calidus, P. phlegethon, A.gilli and juvenile L. capensis in the lower river. A rehabilitation plan is proposed for the Rondegat River. Central to this plan is the formation of a conservancy between the land-owners of the catchment and the custodians of the Cedarberg Wilderness Area. The most effective strategy will be to eradicate M dolomieu from the lower river with piscicides, while at the same time taking steps to rehabilitate the riparian zone throughout the river. A holistic rehabilitation programme such as this would provide an excellent model for future rehabilitation efforts within the CFR.

Includes abstract. / Includes bibliographical references (leaves 52-62). / In the Cape Peninsula, South Africa, the population of chacma baboons (Papio ursinus) has become locally fragmented and geographically isolated from all other populations in the Western Cape. A census performed in 1999 revealed a baboon population under severe human predation pressure with high levels of permanent injury and mortality, in addition to an adult sex ratio strongly biased towards females. There was no data on the causes or the distribution of deaths and permanent injuries on the Peninsula to inform management decisions. The aim of this study was thus to: a) quantify the frequency, distribution and causes of permanent injuries (mutilations) and deaths within the Cape Peninsula baboon population, and b) to quantify whether mutilations, affect the behaviour, diet and reproductive success of baboons.

This research was initiated in 2001 with the overall objective of providing scientific information relevant to the sustainable management of South African Umbrina species. Previously there was confusion over the identity and distribution of Umbrina species in South African waters and very little was known of their life histories, population dynamics and stock status. Morphological comparison of a large number of South African Umbrina (n = 369) with specimens from the type locality (Canary Islands) confirmed the identity of South African U canariensis Valenciennes, 1843, and allowed for an expanded description of the species. However, differences between specimens of U ronchus Valenciennes, 1843 and those of the second South African species, lead to the resurrection of U robinson; Gilchrist and Thompson 1908, as a valid name for this species. Spatial analysis of South African specimens collected with a variety of gear revealed U robinsoni to be a shallow water species found from the surf-zone to 40 m depth that is targeted by recreational shore anglers and spear fishers, whilst U canariensis occurs predominantly from 40-100m depth and is predominately landed as a by-catch in the inshore trawl fishery. Although both species occur throughout the South African eastern seaboard, U canariensis is most common west of the Kei River, where the continental shelf is wider.

A moratorium was placed on fishing for white sharks (Carcharodon carcharias) in South Africa in 1991 prior to knowing the population status of this species. A mathematical model was developed to estimate the total numbers of white sharks from 1950 to 2050 in the presence of fishing (F = 0.055 year-I) and without fishing (moratorium). The total number of white sharks from Richard's Bay to Struis Bay, South Africa was estimated to be 1,954 (range= 1,855-2,050) for the year 2004. If the moratorium was to be lifted and fishing was to resume, the results of the model indicate that the instantaneous fishing mortality rate F of 0.035 year-l or less allows the population to increase in numbers from 2004 to 2050. A sensitivity analysis determined that the instantaneous natural mortality rate M (tested range=0.070-0.l90 year-I) and the average number of pups per mature pupping female (tested range=4-14 pups) were highly sensitive parameters of the model. The maturity age span (maturity minimum age=13 and 16 years; longevity=23, 36, and 60 years) and timeframe of the female reproductive cycle (two or three years) were the least sensitive parameters of the model.

Fieldwork for this study was commenced during the end of January 1969 and terminated in mid-April. 1970. It was conducted on a full time basis and follow-up visits were made to the Park at intervals duririg the remainder of 1970. The study was carried on exclusively in the Bontebok National Park which contains by far the largest herd of bontebok in the country. Visibility in the Park is good and all observations were made from a vehicle with the aid of binoculars at distances ranging from about 30 m to over 400 m. Most animals were tolerant of a motor vehicle but there were one or two exceptions, particularly territorial males whose territories were situated away from the roads and who were therefore unused to being approached closely by vehicles. Females were on the whole more nervous than males and especially females with new born lambs were very nervous and would run from a vehicle at 100 m or more.