Protected areas have been identified as one of the most effective strategies for reducing biodiversity loss in a world where the negative effects of global change are increasing. However for species which migrate or which range beyond the borders of protected areas, these protected areas may only offer partial protection against the threats in the surrounding landscape. Understanding the role and limitations that protected areas can play in conserving threatened species can contribute to better conservation measures for species that may otherwise not benefit from more conventional conservation approaches. The Martial Eagle is a low-density apex predator currently declining across its African range. Changes in reporting rates from bird atlas surveys suggest declines of up to 60 % over the last 20 years (1987-1982 vs. 2007-2013) across South Africa. Worryingly, large protected areas were not immune to these declines. For instance reporting rates in Kruger National Park (KNP; ca. 20,000 km²), an area often considered a stronghold for Martial Eagles, recorded a 54 % decline in reporting rates. It is not clear what the major drivers of declines have been in South Africa, nor what is contributing to the declines in these large protected areas. In this thesis I study the ecology of Martial Eagles in KNP to improve our understanding of the threats they face and how these threats at various stages in their life cycle may be driving declines within protected areas where one would expect that the species should be well conserved. I hypothesised that the main driver of declines in protected areas is that juvenile Martial Eagles disperse beyond the borders of protected areas where they are at increased risk of unnatural mortality, thus leading to recruitment failure back into even the largest protected areas. To test this hypothesis, I fitted GPS tags to 9 juvenile eagles to understand their dispersal behaviour, an aspect of their life cycle for which no previous information existed, and to explore their survival rates. During a lengthy post fledging dependency phase (7 - 9 months) birds began making exploration trips that reached up to ca. 150 km from the nest site and beyond the borders of KNP. After dispersal onset, birds ranged widely up to 390 km from their nests covering areas that averaged ca. 6,500 km²; protected areas covered only 55 % of this area. In contrast to my hypothesis survival rates did not appear particularly low; from monitoring successfully dispersed juveniles over 36 months in total, only one immature bird was confirmed to have died presumably due to natural causes. To understand adult habitat preference and ranging behaviour, which can inform habitat requirements for the species conservation, I fitted GPS tags to eight adult birds. Models of their habitat preference indicated that the species preferred to utilise areas within their home ranges that were in areas with greater tree cover, with areas of dense bush rather than open bush or grassland, amongst other important features. These results were important to identify potential threats, such as the loss of trees in Savannah's, which is currently occurring due to elephant damage and fire influences. The species held large territories (ca. 108 km²) constraining the maximum number of pairs that the Park is able to support (max. 185 possible pairs), however models of distribution suggest the available habitat in KNP likely supports ca. 60 - 70 breeding pairs. Two adult individuals never held territories and another two abandoned their territories during the course of the study. These individuals ranged widely (ca. 44,000 km²) suggesting a floater population exists in the region. The death of three of these four floater individuals (two persecutions and an electrocution) indicates that adults are particularly at risk of mortality during these wide-ranging movements beyond protected area boundaries. Two natural mortalities of territorial birds within the park were also recorded. Overall therefore, despite our relatively small sample size, adult birds do appear to have worryingly low survival rates. Because my sample size of both adults and juveniles/immatures was relatively small (adults = 8, juveniles = 9), their movements may not be fully representative of the entire population. Therefore, I additionally modelled the distribution of Martial Eagles using independent sightings data to describe suitable areas for the species both within the KNP and adjacent areas (within ca. 400 km of the park). Identifying these areas provides conservation managers with more information to ensure adequate conservation measures are in place for this species in these areas. At least 29 % of KNP was predicted to be suitable for Martial Eagles, while neighbouring regions in Mozambique and Swaziland were also predicted to be highly suitable for the species. Given the adult mortalities and general scarcity of Martial Eagles in Mozambique the area may act as a population sink for KNP birds. Lastly, I compared current reproductive parameters to those reported in a considerable number of other studies on the species both within KNP and elsewhere. Productivity recorded during this study was lower than any previously recorded estimate. Using a population model, I show that current productivity within KNP is sufficiently low to have been solely responsible for the known levels of decline there, without the need to invoke any other contributory factors. A high hatching failure rate was mostly responsible for the low productivity. However, it is important to note that at least two of the three years of data collection occurred during low (drought) rainfall years, which may have constrained breeding, and thus may not be reflective of productivity levels more generally over the recent longer term. My research helps identify the most likely drivers of population declines in KNP, suggesting that elevated adult mortality and lower productivity may be the key factors. Drivers of low productivity require further investigation, however it is likely that changes in habitat quality or climate may be impacting on the species within KNP. The study also highlighted the difficulty of conserving wideranging and threatened species in protected areas, which may be prone to high mortality in the surrounding landscape. This research is therefore applicable to a number of species that range widely from KNP e.g. vultures, or migratory eagles. The research indicates that protected areas alone are unlikely to conserve these species and that additional conservation measures, such as education programmes, or trans boundary policy should be put in place to realise successful conservation for these species.

A general and increasing biodiversity loss has been observed since the 20th century. Faced with the extreme rapidity of population declines, conservation biologists seek to understand the limiting and regulating factors driving changes in animal populations. This is particularly important for rare species as small population size increases extinction risk. Birds are amongst the most studied animals in this context. As a group that occupies a high trophic level, raptors are particularly vulnerable to external changes and are generally regarded as useful indicators of ecological change. The Black Harrier Circus maurus is an avian predator endemic to southern Africa, which breeds essentially along the South African coast within the Fynbos biome, and inland within the Karoo biome. Its population size has been estimated at less than 1,000 breeding birds, and the species is currently considered as Endangered in South Africa, Namibia and Lesotho. Although some studies have been conducted on Black Harriers in the last four decades, the reasons for its scarcity currently remain little known and insufficiently explored. Filling this knowledge gap is therefore essential for its conservation. In this context, the main goal of this thesis is to develop an overall comprehension of how various environmental factors may affect the breeding and health of this Endangered species, at both population and individual levels. I conducted my fieldwork during the 2012-2015 breeding seasons in two contrasting geographical regions: one along the west coast in the Western Cape Province, and the second one inland in the surroundings of Nieuwoudtville in the Northern Cape Province. For some chapters (Chapters 1-3), I analysed historical data collected by Dr. R. E. Simmons during 2000-2011 breeding seasons.

A review of the genus Cisticola was published in 1930 by Rear-admiral Lynes. While subsequent authors have modified Lynes' original groupings, his work remains the basis for modern syntheses of cisticolas. This study tests Lynes' hypotheses by analysing data that he presented in his review and with measurement and plumage data collected from museum specimens. Lynes' groupings were well recovered (98%) when data captured from his review were analysed phenetically, suggesting that he grouped species mostly by similarity. In contrast, when morpho-behavioural data were analysed using cladistic methods, many of his groupings were not monophyletic and the resultant cladogram had very little nodal support due to their highly conservative morphology. To resolve the structure of the genus and the relationships within it, two mitochondrial and four nuclear regions were sequenced from toe-pad samples taken from museum specimens. The molecular analyses included 44 of the 49 currently recognised species and represents the most taxon-dense molecular phylogeny of the genus. The resultant phylogeny separates species into five main clades, but many of Lynes' groupings were not monophyletic and there was also very little support for more recent groupings. Vocalisation analyses indicated that frequency components of songs were correlated with habitat type and body size. These correlations, though, disappeared when phylogeny was controlled for indicating that phylogenetic history rather than habitat preference influenced song character distribution. Some song types are mismatched to their environment, and some sympatric sister species appear to give similar calls. Cisticolas may overcome these attenuation and identification difficulties with behavioural adaptations and aerial displays. The biogeographic distribution of closely related species does not agree with many of the previously proposed hypotheses and a dated phylogeny estimates that most of the diversification in the genus has occurred within the last five million years. Most of the mean divergence date estimates correlated with periods of climate variability and episodes during which there is evidence for high lake levels in Africa, rather than correlating with Plio-Pleistocene glaciation, offering evidence that open habitats may have become fragmented during extremes of both arid and humid climates.

Their high energetic demands make seabirds sensitive to changes in prey availability, which is often reflected in their diet and energetic expenditure during breeding. Populations of the three seabirds endemic to southern Africa's Benguela upwelling ecosystem that rely on small pelagic fish have decreased dramatically over the last decade. In contrast, the population of the greater crested tern Thalasseus bergii has increased. To understand these conflicting trends, I investigated the foraging ecology and energy requirements of greater crested tern breeding in the Western Cape, South Africa. Diet was assessed by a novel non-invasive methodology developed in this study, using digital photography. More than 24,000 prey items from at least 51 different prey taxa were identified, with 34 new prey species recorded, revealing a high degree of foraging plasticity for this seabird. Greater crested terns rely mainly on anchovy Engraulis encrasicolus (65%), which averaged 84 mm long. Prey composition differed significantly between breeding stages, with anchovy especially dominant at the onset of the breeding period and the diet becoming more variable as the season progressed. Time-energy models for breeding terns were built based on activity budgets collected from non-invasive video-recordings and focal observations. Foraging trips were significantly longer during incubation than the chick provisioning stages, and feeding rates doubled from early to late chick provisioning. This study illustrated a steady increase in energy needs of adults throughout the breeding season, due to their increased foraging effort to meet chick energy needs. In comparison to other Benguela endemic seabirds that also rely on small pelagic fish, terns displayed substantially lower energy requirements at both individual and population levels. I also explored the benefits underlying interactions within mixed-species aggregations by investigating the costs induced by kleptoparasitism between mixed colonies of greater crested terns and Hartlaub's gulls Chroicocephalus hartlaubii and colonies with greater crested terns alone. Video-recordings coupled with focal observations showed that terns suffer direct costs to chick provisioning rates and indirect costs through energy expenditure in a mixed-species colony, suggesting that these breeding assemblages may be a form of parasitism rather than a mutualistic association. Despite the detrimental effects of interspecific kleptoparasitism, the marked foraging plasticity and low energetic requirements of greater crested terns, described in this study, coupled with specific life history traits such as low fidelity to breeding sites and extended post-fledging care, are key factors that allow this species to cope with changes in the availability and abundance of their main prey. Understanding species-specific behavioural responses to ecosystem variations in the Benguela upwelling system is vital for assessing the impact of commercial fisheries on seabird populations and fish stocks. Finally, the implementation of the method developed in this study, in long-term monitoring programmes, may provide crucial knowledge for conservation plans and key input to realising an ecosystem approach to fisheries management.

Seabirds are considered one of the most threatened groups of birds in the world. They face additional mortality both on their breeding islands from introduced predators and at sea by fishing fleets, as fisheries bycatch, as well as other human impacts. Seabird bycatch has negatively affected many seabird populations worldwide, with trawl, gillnet and longline fisheries considered the most destructive to seabird populations. Seabird bycatch from trawl and gillnet fisheries has been significantly reduced in recent years, but large numbers of seabirds are still killed annually by longline fisheries. Of the two types of longline fisheries (demersal and pelagic), pelagic longlining is considered the most harmful to seabirds as lines remain closer to the surface for longer periods than demersal longlining, and it is harder to weight lines to ensure rapid sinking beyond the depth they are accessible to birds. Seabirds are killed when they swallow baited hooks and consequently drown. Despite the large number of studies investigating seabird bycatch from pelagic longline fisheries, there remain gaps in our understanding of seabird bycatch from pelagic longline fisheries. This thesis addresses some of these knowledge gaps and makes recommendations as to how seabird bycatch from pelagic longliners can be reduced at both a local and global scale. Chapters 2 and 3 investigate the factors affecting seabird bycatch from pelagic longliners off South Africa, provide a summary of seabird bycatch from the fishery for the period 2006–2013 and quantify the structure of seabird assemblages associated with pelagic longline vessels off South Africa. This was achieved by analysing seabird bycatch data collected by fisheries observers as well as data from sea trials onboard pelagic longliners. Seabird bycatch by pelagic longliners off South Africa over the 8-year study period has been significantly reduced from the 8-year period (1998–2005), mainly driven by a significant reduction in seabird bycatch rates from foreign-flagged vessels, which are responsible for c. 80% of fishing effort off South Africa. Seabird bycatch rates from South African vessels still remain high, four times higher than the interim national target of < 0.05 birds per 1000 hooks. The species composition of seabird bycatch off South Africa is best explained by an understanding of the structure of the seabird assemblage associated with longline vessels. For most species, bycatch and attendance ratios were similar, but for some species such as shy-type and black-browed albatrosses there were large mismatches, likely caused by differences in foraging behaviour and foraging dominance hierarchies. In Chapters 4 and 5 the foraging ecology of the most commonly recorded bycatch species off South Africa, the white-chinned petrel (Procellaria aequinoctialis), is investigated. An understanding of the foraging ecology of commonly recorded bycatch species enhances our understanding of seabird bycatch and helps to improve the design of current and future mitigation measures. The year-round movements of white-chinned petrels from Marion Island were investigated with Global Location Sensors (GLS loggers) and GPS loggers. Adult white-chinned petrels undertake only limited east-west movements of, with all birds remaining between southern Africa and Antarctica. These results strengthen the theory that there is limited spatial overlap year-round between white-chinned petrel populations from South Georgia, the southern Indian Ocean islands and New Zealand sub-Antarctic islands, suggesting that these populations can be managed as separate stocks. The diving behaviour of white-chinned and grey petrels (P. cinerea), another common bycatch species in Southern Ocean longline fleets, were examined with the use of temperature-depth recorders (TDRs), deployed on birds from Marion Island and Gough Island. My study was the first to analyse TDR dive data from any Procellaria petrel, and recorded them reaching maximum dive depths of 16 and 22 m, respectively. Current best practise suggests that baited hooks be protected to a depth of 5 m by bird-scaring lines, but my results suggest this depth should be increased to at least 10 m. Although line weighting is a proven mitigation measure to reduce seabird bycatch from pelagic longliners, fishers have concerns that it will compromise fish catches, crew safety and operational efficiency. In Chapter 6 I analyse line weighting data from trials onboard three pelagic longline vessels, to address the concerns of fishermen. My results show that sliding leads can be incorporated into pelagic longline fisheries without compromising fish catch, crew safety or operational efficiency. I thus recommend that sliding leads be used on pelagic longline vessels fishing off South Africa. By incorporating studies investigating the factors affecting seabird bycatch, seabird foraging ecology and the efficacy of seabird bycatch mitigation measures, my thesis has broadened our understating of seabird bycatch from pelagic longliners and makes meaningful recommendations to further reduce bycatch, both locally and globally. Although seabird bycatch rates have declined off South Africa, through the use of a number different mitigation measures, they still remain higher than the South African national target and thus more work is needed to achieve this target. To reduce seabird bycatch from pelagic longliners to acceptable levels, studies from the world's various longline fleets needs to be considered and improved upon, with seabird conservationists and fishermen working together to achieve this goal.

Populations of secondary tree-cavity nesting bird species are often limited by a shortage of natural nesting sites. For the Southern Ground Hornbill Bucorvus leadbeateri that typically nests in natural tree cavities, the shortage of nesting sites is one factor potentially limiting population growth. The species is listed as endangered in South Africa, and vulnerable throughout the rest of its range. Nest boxes can improve the conservation status of threatened birds that are limited by nest-site availability. However, nest boxes or other types of artificial nests are not always beneficial to the target species, and their value as a conservation tool needs to be tested for each species. Wooden nest boxes were installed for ground hornbills in a study area in north eastern South Africa with a paucity of natural nest sites. In this thesis, I assess productivity, timing of breeding, and dispersal in the Southern Ground Hornbill in a study area supplemented with nest boxes and discuss the implications for the conservation of this endangered species. Nest boxes are an effective conservation tool to improve productivity in areas lacking natural tree cavity nesting sites. Breeding success (calculated as the proportion of nesting attempts that fledged a chick) and predation levels were similar for groups using nest boxes and natural nests. Natural nests were more buffered against cooling night temperatures, but otherwise nest boxes provided nesting conditions that were no better than natural nests. Timing of breeding for nests in natural tree cavities and nest boxes were similar. However, groups with access to a nest box attempted breeding more often than groups with access to a natural nest only, resulting in an 15 % increase in the number of fledglings per group compared to an adjacent protected area with no artificial nests. The number of breeding groups in the study area increased by 460 % over 12 years. However, there is a limit to the density of breeding groups. Breeding success was highest when breeding density was one breeding group per 90-120 km², so nests should be spaced ~10 km apart. Given that the threats to ground hornbills include persecution and poisoning, increasing the reproductive rate by providing nest boxes should assist in slowing the decline by the increased recruitment of offspring into the population. Timing of breeding varied across years. The first eggs laid each year ranged from 9 September to 14 November, and median lay date was 03 November. Breeding attempts that were initiated early in the season were more likely to fledge a chick than those initiated later in the season. Timing of breeding was delayed during warmer springs, particularly under dry conditions. In savannas, hotter spring temperatures could limit food availability, for example, if higher temperatures cause the vegetation to dry out, resulting in a rapid decline in insect flush, especially in the phytophagous insect groups that form a large part of the ground hornbill diet. Factors to consider when constructing and placing nest boxes include thickness of the cavity walls, entrance height above ground and density of nest boxes placed in the landscape. Breeding attempts in natural nests and nest boxes with thicker nest walls and those positioned with higher entrances above the ground increased breeding success. Therefore, nests should be constructed with cavity walls at least 6 cm thick and placed so that the entrances are situated > 6 m above the ground. With 186 ringed chicks fledging from the study area after the installation of nest boxes, it was possible to observe their dispersal within the study area and farther away into the adjacent Kruger National Park. There was no evidence for sex-biased dispersal. Males and females dispersed at similar ages, and over similar distances, raising interesting questions about inbreeding avoidance mechanisms in this species. If females do not disperse beyond the range of related males, how do related individuals avoid pairing, and what forms of individual recognition exist?

The effects of environmental warming on the reproductive performance of birds are most easily studied in desert habitats where birds already experience air temperatures (Tₐs) close to their upper thermal tolerance. Many desert birds coincide breeding with periods of food availability triggered by rainfall during the summer season. Daily maximum air temperatures (ₘₐₓ) during the Kalahari summer season frequently reach the lower forties (°C) and recent years have been characterised by reduced rainfall and increased Tₐ. Breeding Southern Yellow-billed Hornbills (Tockus leucomelas) could be particularly vulnerable to high Tₐ due to their breeding strategy whereby the females are confined to the nest cavity for most of the nesting period. During this time their male partners are solely responsible for food provisioning, which imposes a considerable energetic demand. In this thesis, I investigated the extent to which Tₐ affects the ability and willingness of breeding males to provision their female partners and offspring. And consequently, the extent to which male investment and the thermal environment affect female body mass (Mb) and chick development rates in Southern Yellowbilled Hornbills in the Kalahari. During three consecutive hornbill breeding seasons (October - March, between 2012 and 2015), I collected life history data during 50 breeding attempts by 32 hornbill pairs. At the study site, Southern Yellow-billed Hornbills readily breed in artificial nest boxes and this allowed me to assess the internal nest climate using temperature and relative humidity loggers which were placed in most of the nests. The male hornbills in the study population were semi-habituated which facilitated behavioural observations. Weather data were recorded at an on-site weather station. Morphometric data from females and chicks were collected on a daily basis at selected nests and perch scales installed at nest entrances recorded Mb data of the provisioning males. From chick hatching to chick fledging, I observed the behaviour of the males during 30-min focal follows and focussed on foraging behaviour, prey allocation decisions (nest versus self), microsite use and thermoregulatory behaviour. Male hornbills spent more than half of their time panting at Tₐs above 34.5 °C. Days on which this threshold temperature was exceeded were therefore described as 'hot days'. The male hornbills experienced trade-offs on hot days between foraging efficiency and panting behaviour, indicating that the additional cost of thermoregulation and high Tₐ affected foraging success (Chapter 2). Males would always provision larger prey items to the nest and consumed the smaller prey items themselves. As Tₐ increased, the males increased their foraging effort, but caught fewer and smaller prey items overall, reducing the total biomass they provisioned to the nest as well as the biomass they consumed. As a result, males were unable to maintain their Mb on days when Tₐ exceeded 37.9 °C (Chapter 3). A similar effect of hot days on Mb maintenance was observed in females and chicks within the nest. Independent of chick age, females departed the nest when their Mb reached a lower limit of 189.3 ± SD 18.1 g. The females would then aid the males in nest provisioning, however the negative effect of increasing ₘₐₓ on provisioning rate was still evident; i.e. females were not able to compensate for reduced male provisioning rates on hot days. High Tₐs during the nesting period resulted in smaller and lighter fledglings and overall reduced the probability of a successful nesting attempt (Chapter 4). A thermal imaging experiment revealed that the large beak of hornbills (both males and females) plays an important role in non-evaporative heat loss. Hornbills were observed to dissipate up to 19.9 % of the total non-evaporative body heat loss via the beak. This water-saving mechanism can be highly advantageous to hornbills living in arid regions where water availability is limited (Chapter 5). Lastly, a comparison of the results of the current study with those of a study on the same hornbill population carried out between 2008 and 2011 revealed that mean ₘₐₓ as well as rainfall during the nesting period had an important impact on overall hornbill reproductive effort and success (Chapter 6). Long-lived species are expected to prioritise future reproductive opportunities over current broods. However, the predicted scenario for the Kalahari is that high Tₐs become more extreme and periods of drought become more frequent. Therefore, I predict an increased risk of breeding failure among Southern Yellow-billed Hornbills in the future which could affect the persistence of this population.

The habitat of an animal is extremely important as it provides that animal with the necessary resources for fulfilling its life-history requirements (Brennan &amp; Block, 1993; Beyer et al. 2010). A habitat is defined as a region in environmental space which comprises of multiple abiotic and biotic variables influencing an animal's location (Krausman, 1999; Beyer et al. 2010). Animals tend to utilise discrete areas within a habitat, constituting part of their home range. Home range analysis helps to delineate the area used by an animal habitually and areas of concentrated use (Samuel et al. 1985; Seaman &amp; Powell, 1996; Moorcroft et al. 1999; Mitchell, 2007; Rodgers &amp; Kie, 2011). Patterns of differential use of space within an animal's home range are the result of competing demands and trade-offs. In this study, a kernel technique was used to determine the home range of four satellitetracked groups of Southern Ground-Hornbills Bucorvus leadbeateri in the Associated Private Nature Reserves (APNR) in the South African lowveld. Satellite data were analysed in ArcGIS® 9.3 to quantify habitat selectivity by groups of ground-hornbills at different times of the year to determine a) favoured habitat types, and b) the resolution with which they perceive their environment. Each of the I groups showed variation in the utilisation and extent of their home ranges on a seasonal basis. Home range sizes contracted towards the nest during the summer breeding season (December to March) and expanded during the dry season (April to September). Within the home range of one of the groups the physical characteristics of habitat types (i.e. vegetation types) were sampled at 250 random co-ordinates in order to assess whether habitat preference at the meso-scale can be explained by the physical attributes of that vegetation type. By profiling and quantifying the vegetation of areas in the home range that are used by Southern Ground-Hornbills to differing degrees, this information could be used as a proxy to facilitate re-introduction efforts, by providing a tool to identify optimal landscape configurations.

Includes bibliography. / A study of the ecology of the Southern African Kestrel was made in two areas of differing topography in the Eastern Cape Province (32 °S) of South Africa. The main objectives of the study were to describe foraging parameters (behaviour-time budgets, hunting success and diet) and breeding parameters (clutch size, hatching and fledging success, nestling, fledging and post-fledging periods) of the Southern African Kestrel. These parameters were compared to those from studies made on the European Kestrel (Falco tinnunculus tinnunculus) in western Europe (53 °N) in an environment which fluctuates relatively more than the South African study sites. Predictions concerning foraging and breeding ecology made by life-history theory in stable and fluctuating environments prompted the erection of the following hypotheses which were examined in this study. 1. Common Kestrels (Falco tinnunculus) in relatively stable environments have a more consistent but lower foraging yield than those breeding in fluctuating environments. 2. Common Kestrels in relatively stable environments have lower reproductive rates (longer incubation, nestling and post-fledgling dependence periods, later age at first breeding and longer life-span) than those in fluctuating environments. Southern African Kestrels spent different proportions of time of the two main hunting methods used, perch- and hover-hunting, than the European Kestrel. Perch-hunting was the predominant hunting method used by the Southern African Kestrel compared to hover-hunting in the European Kestrel. Southern African Kestrels had a higher prey-capture success for both hunting techniques. However, daily food intake was lower in Southern African Kestrels because a high proportion of the diet consisted of small invertebrates. Average clutch size was smaller in the Southern African Kestrel than that of the European Kestrel and Southern African Kestrels fledged on average fewer young. However, incubation, nestling and post-fledging periods were within the range recorded for the European Kestrel. The hypothesis that Common Kestrels in relatively stable environments have lower foraging yields is supported by this study. Although results from this study suggest that food yields are more consistent in stable environments, it was unable to evaluate this adequately. The hypothesis that reproductive rates are lower in relatively stable environments is partly supported in this study, but more data are required for parameters such as incubation and nestling periods to confirm such trends.

The African penguin Spheniscus demersus is a seabird endemic to the south-western coast of Africa and can be found in three main breeding localities; southern Namibia, the Western Cape and Algoa Bay, Eastern Cape, South Africa. The African penguin has been listed as Endangered since 2010, having experienced a decline in population of more than 50% over the past three generations. This study was conducted at the breeding colony on Robben Island, South Africa, and examined two factors that have the potential to affect breeding success of African penguins; body mass at the onset of breeding and the suitability of various nest types to mitigate changing climatic conditions. Body mass was measured by recording weights using an automated weighing scale set up in front of a nest. Weights were taken at the start of breeding of each penguin in a breeding pair and these weights were compared to the number of chicks fledged, fledging period, hatching success, clutch size, and chick fledging weight. Hatching success, clutch size, and fledging weight were not influenced by the mass of either parent. There was a trend of shorter fledging periods as the mass of the heavier parent increased. The greatest effect was from the body mass of the lighter parent on the number of chicks that fledged from the nest; as the mass of the lightest adult increased, more chicks were fledged. If the lighter adult weighed below 2 kg there were always no chicks at the nest that fledged. This suggests evidence for a carry-over effect of body mass from the time before breeding starts into the breeding season, and highlights the importance of food availability for African penguins on a global scale, and not just a local one. The proportion of nest failures of six nest types (vegetation, open, natural burrow, building, wooden nest box, and artificial fibreglass burrow) at the incubation and chick-rearing stage were compared to rainfall and maximum temperature.