Mammalian parental investment (i.e. care of descendant offspring) is largely biased towards maternal contributions due to the specific feeding needs of mammalian offspring; however, varying degrees of paternal investment have been reported in about 10% of all mammalian species. Within the order Carnivora, paternal contribution to rearing offspring is particularly high: an estimated 32% of all studied carnivore species exhibit direct paternal care. Despite the prominence of paternal investment in carnivores, the endocrine basis of this behaviour is not well understood. This review examines the current — highly constrained — state of knowledge about the endocrine basis of carnivore paternal investment. We attempt to link changes in androgen and glucocorticoid levels with variation in direct and indirect paternal care behaviour making specific predictions regarding the way forward. Well-studied species, such as bat-eared foxes (Otocyon megalotis), dwarf mongoose (Helogale parvula) and meerkats (Suricata suricatta), where social dynamics are relatively well understood, can act as ideal model systems through which we may further investigate the endocrine basis of paternal investment in carnivores.

Several national parks on the African continent have been fenced to mitigate human-wildlife conflict, but the effectiveness of different fencing strategies seems to vary between regions. Here, we assessed the effectiveness of electro-fencing for restricting movement patterns of several wild ungulates in Akagera National Park (NP), Rwanda. Following the turmoil of the Rwandan genocide and civil war (1991–1995), large numbers of returning war refugees brought about an increasing pressure on land use, after which considerable parts of the western Akagera NP and adjacent Mutara Game Reserve were degazetted, followed by the construction of an electric fence in 2013 and a helicopter-based game-drive of ungulates into the modern park. We conducted distance sampling of wildlife and domestic livestock in the Uruhita plain — now no longer part of Akagera NP but bordering the modern Akagera NP to the west — for extended periods of time (2010–2015) before and after fencing. For abundant species population densities could be calculated, whereas for other species encounter frequencies are reported. We report on continuously high cattle densities, and increasing encounter frequencies of sheep/goats in the Uruhita plain. We found decreasing densities of zebra (Equus quagga) and decreasing encounter frequencies of waterbuck (Kobus ellipsiprymnus), stagnant encounter frequencies in the case of topi (Damaliscus lunatus jimela) and stagnant densities of impala (Aepyceros melampus). The latter species can jump fences, and we argue that competitive release and reduced predation may be the prime factors leading impala to (temporarily) exploit grasslands outside the modern national park.

Breeding success (fledglings pair^-1 y^-1) of the Red-listed African Black Oystercatcher (Haematopus moquini) is highly variable, both spatially and temporally. Despite a diversity of natural factors causing this variability, there is evidence that two anthropogenic factors, i.e. disturbance and an introduced mussel (Mytilus galloprovincialis), are having an impact on the local breeding success of this species. Using a data set comprising 87 site-years of nest-monitoring data across most of the species' breeding range, we analysed the extent and causes of variability in breeding success. Breeding success differed across three population categories defined by varying levels of human disturbance: island populations, protected mainland populations, and unprotected mainland populations. Differences in breeding success between island populations and protected mainland populations were likely due to differing exposure to predators; however, differences between protected and unprotected mainland populations were unlikely caused by this as both experience equivalent predation levels (although from different predators). Protection only improved the breeding success of oystercatchers in very high-quality habitats (with a high biomass of alien mussels), and where populations were ‘released’ from high levels of human disturbance. In unprotected mainland areas, human activity impacted on the breeding success of local populations primarily through predation of small chicks by uncontrolled dogs, and by rising tides drowning chicks that were hiding from human disturbance. The findings of this study note the potential conservation dilemma resulting from an invasive species improving the conservation status of a Red-listed species, and encourage the implementation of restricted sites in high-quality habitats with high breeding pair densities.

Two introduced beachfleas, Orchestia gammarellus and Platorchestia platensis, have been recorded from South Africa, but have been misidentified by some earlier researchers and, prior to this study, each was also known from just a few, scattered observations, such that their true distributions remained obscure. We illustrate both species to clarify the features that identify and distinguish them and determine their true distributions by re-examining historical samples and by collecting new material from estuaries and lagoons along the west and south coasts of South Africa. Adult males of O. gammarellus have slender antennae and characteristic expanded, flattened, oar-like distal segments on pereopod 7. Adult males of P. platensis have the peduncle of antenna 2 strongly swollen and have bulbous, but not flattened, distal segments on pereopod 7. Both species occur under wrack and other debris along the banks of estuaries and sheltered lagoons. Orchestia gammarellus is found only along the south-west coast, in the Berg River Estuary, Langebaan Lagoon and the Diep River Estuary. Previous records from Knysna were misidentifications of P. platensis. Platorchestia platensis is found at one west coast site, Langebaan Lagoon, but is widely distributed from Cape Point eastwards to at least Algoa Bay. Both species share the same habitats as several other native beachfleas and competitive interactions between native and introduced species merit investigation.

Rhinoceros populations in Africa are under severe threat as a result of surging poaching rates and risk-mitigation strategies are continuously adapted in an attempt to ensure the survival of the species. This study compared faecal glucocorticoid metabolite (fGCM) levels of two age classes of limited free-ranging female white rhinos with fGCM levels of adult free-ranging female white rhinos. Subsequently, fGCM alterations in the limited free-ranging animals were monitored following routine dehorning as a measure of the animals' short-term physiological stress response. Baseline fGCM levels differed significantly between tested groups, with both free-ranging and limited free-ranging adult animals showing significantly higher fGCM levels compared with limited free-ranging juvenile females. In contrast, baseline fGCM levels did not differ significantly between limited free-ranging and free-ranging adult individuals. Routine dehorning procedures resulted in a short-term stress response expressed by a significant increase in fGCM levels 48 h post-dehorning, with stress steroid levels returning to pre-dehorning concentrations 72 h after the procedure.

A number of predator-prey studies show that certain prey demographic classes are targeted over others. The possible mechanisms driving these effects can be two-fold. Firstly, a preference for a certain demographic class results in selective predation by a predator. Secondly, different demographic classes exhibit varied behaviour and thus susceptibility to predation risk. To test these mechanisms, a study was conducted on Samara Private Game Reserve to investigate the potential impact cheetah (Acinonyx jubatus) predation has had on the kudu (Tragelaphus strepciseros) population. Kudu age and sex data were collected across both predator-present and predator-absent sections using a spotlight-count method. Results suggest that juvenile kudu and subadult male kudu are selectively hunted by cheetah. This was reflected in significantly lower proportions of these two demographic classes within the predator-present section (both 6.0%, respectively) compared with the predator-absent section (16.8% and 17.4%, respectively). These findings indicate that the shift in the demographics of the kudu population within Samara could be explained by a combination of these two mechanisms.