Bees are herbivorous insects, consuming nectar and pollen throughout their life cycles. This paper is a brief review of the chemistry of these two floral resources and the implications for bee nutrition. Nectar is primarily an energy source, but in addition to sugars contains various minor constituents that may, directly or indirectly, have nutritional significance. Pollen provides bees with the protein, lipids, vitamins and minerals that are essential for larval rearing. Chemical analyses of pollen have tended to focus on the protein component of bee-collected pollens as an index of nutritional quality. However, the substantial nectar content of such samples (∼ 50% dry mass) should not be ignored, especially in view of current interest in measuring the nutritional quality of floral resources for bees.

There is a paucity of studies examining direct impacts of introduced alien species on biodiversity, a key need for motivating for alien species control in conservation areas. The introduced prickly pear (Opuntia stricta) has invaded some 35 000 ha of Kruger National Park. We investigated the effect of O. stricta on beetle and spider species assemblages in the Skukuza region of Kruger National Park. We used unbaited pitfall traps over a 12-month period in four treatments of varying O. stricta density. Species richness, species density and abundance of beetles and spiders were compared. A total of 72 beetle and 128 spider species were collected. Species richness and species density for beetles and spiders did not differ significantly across the four treatments. Assemblages for spiders did not differ across treatments but beetle assemblages were significantly different from uninvaded control sites. This study suggests that the current density of O. stricta does not significantly affect spider species richness, density or assemblages but that beetle assemblages are significantly affected.

The trophic structure and biomass of macrobenthos on both wave-sheltered, rocky intertidal shores and semi-exposed ones at seven localities in the Tsitsikamma Marine Protected Area were compared. In the Cochlear zone and entire intertidal (Cochlear- and Balanoid zones combined) biomass values of invertebrates and filter-feeders were significantly (P < 0.05) higher at semi-exposed shores, while grazer biomass was greater at sheltered sites: a negative relationship (P < 0.01) and a moderate to weak correlation (r > 0.42) was also recorded between filter-feeder biomass and grazer biomass. This negative relationship reflected changes in standing biomass of the two dominant filter-feeders (Octomeris angulosa and Perna perna) against that of the dominant grazer (Scutellastra cochlear). On sheltered shores the biomass of trophic groups varied significantly (P < 0.05) across the intertidal gradient studied, with values of grazers and algae decreasing with increased aerial exposure. Similar, but less substantial (P > 0.05) variations were recorded at semi-exposed sites, where the median shell length of the brown mussel P. perna decreased upshore (P < 0.05). Multivariate analyses of square-root transformed biomass values separated most of the sheltered Cochlear zone stations from semi-exposed ones. This study supported the concept that the structure of low-shore communities is determined largely by differences in wave action.

The effects of timing, frequency and duration of mouth opening and marine overwash events on fish assemblages in two, adjacent temporarily open/closed estuaries were investigated. Sampling was conducted biannually over two years during summer and winter, using selective sampling gear to target estuarine-associated marine species. During this period the East Kleinemonde Estuary opened more frequently than the West Kleinemonde Estuary and it also experienced a greater number of bar overwash events when larval recruitment can occur. The timing of mouth opening and the frequency of marine overwash events did not influence species richness of marine-spawning species but had a pronounced effect on abundances (catch-per-unit-effort) of species. The species composition of both estuaries was similar, but the catch-per-unit-effort in the East Kleinemonde Estuary was more than three times greater than that of the West Kleinemonde Estuary. The East Kleinemonde also supported a greater proportion of immature individuals of Rhabdosargus holubi and Monodactylus falciformis, and the length frequency distributions of both these species differed significantly between the two estuaries. The larger fish sizes in the West Kleinemonde Estuary were attributed to the prolonged mouth closure that resulted in the trapping of larger juveniles and mature fish within the estuary.

The taxonomy of the shrews (family Soricidae) is primarily underpinned by the analysis of morphology characters. These morphological characters are, however, often plesiomorphic and difficult to interpret in an evolutionarily meaningful way. Recent molecular studies have provided some clarity at lower-taxonomic levels, but studies giving resolution to deeper levels of the soricid phylogeny are lacking. By constructing the most taxonomically inclusive molecular phylogeny for this cosmopolitan family, this study aims to test the utility of the routinely used mitochondrial cytochrome b gene to resolve the higher-level relationships within the Soricidae. Given the life history characteristics of shrews (high metabolic rate in Soricinae, small body size, fast generation time and generally short life span), the metabolic theory predicts accelerated nucleotide turnover in this family. By examining the molecular evolution of the cytochrome b molecule, this study identifies both lineage- and codon-specific mutational rate differences. We suggest that codon-specific mutational rate patterns are best explained by unique selective regimes operating on the different codon positions, while lineage-specific rate variation among genera belonging to Soricidae may reflect differences in the mode and speed of speciation events which generated species.

Spatial patterns in the macrofauna community structure within four distinct zones of the permanently open, freshwater-deprived Kariega Estuary on the southeastern coastline of southern Africa were investigated in March–May 2010. The zones within the littoral zone comprised a band of Zostera capensis (Setchell) exposed at spring low tide and comprising Zone I; a band of mud or sand, lacking vegetation, corresponded to Zone II; stands of Spartina maritima (Curtis) formed Zone HI and a belt of Sarcocornia perennis (Miller) constituted Zone IV Total macrofauna abundance and biomass during the study ranged from 16–816 ind/m² and from 0.032–390.76 mg wwt/m² and demonstrated no significant horizontal or vertical patterns (P > 0.05 in both cases). Multidimensional scaling (MDS) indicated that there were no significant spatial patterns in the macrofaunal community structure within the four zones which could be related to the predominance of euryhaline species, including Marphysa sanguinea (estuarine wonder worm), Arcuatula capensis (estuarine mussel), Macoma litoralis (littoral tellin) and Nassarius kraussianus (tick shell) throughout the estuary. The absence of any distinct spatial patterns in the community structure of the macrofauna appears to be related to the virtual absence of any horizontal patterns in selected physico-chemical variables and availability of submerged macrophytes along the length of the estuary. Comparisons with historical data indicate that while there have been no significant changes in the estimates of the total macrofauna abundance and biomass within the estuary over the past three decades, abundances of selected species have declined by as much as 80%. The decline in the abundances of these species appears to reflect the predominance of hypersaline conditions within the upper reaches of the estuary and intense fishing activities in the lower reaches of the system.

Allometric growth patterns of functional morphological variables that reflect resource use in fish were correlated with the diet of Diplodus capensis in southern Angola. A total of 114 individuals (76–336 mm FL) were collected for stomach content analysis and a further 35 individuals (80–320 mm FL) for morphometric analysis. The major dietary components belonged to the groups Chlorophyta, Cirripedia, Rhodophyta and Bivalvia, and made up 71.7% of the diet in terms of a computed ranking index. There was a clear ontogenetic dietary shift with small fish feeding predominantly on algae and large fish on barnacles and mussels. This increasingly durophagous diet corresponded with a positive allometric growth pattern in the gape dimensions (width, height and surface area), adductor mandibulae muscle process and incisor width. Conversely, the relative gut length followed an isometric growth pattern, suggesting a decreased absorptive capacity. This isometric growth pattern is attributed to the better digestibility of the durophagous prey items when compared with algae and further corroborates a link between the morphology and an ontogenetic dietary shift observed in this species.

Although urbanization is perhaps the most damaging, persistent, and rapidly expanding form of anthropogenic pressure on natural ecosystems, data on the patterns and processes of sympatric bat species in urban landscapes are relatively scant. We quantified the packing and dispersion of sympatric animalivorous bats based on flight and echolocation parameters at two urban rivers in Durban, South Africa. We used null models to test if the observed phenotypic patterns deviated significantly from the random patterns expected in the absence of competition or prey defences at ensemble and functional group scales. As we predicted, species packing increased in the species-rich ensemble comprising many morphologically similar species that grouped together, with a few morphologically distinct species dispersed further away. Furthermore, we found evidence that competition influenced flight patterns of the open-air bats at the species-rich river ensemble, and prey defences influenced echolocation patterns of coexisting bats at the ensemble and functional group scales. However, non-deterministic processes such as habitat structure and the physics of sound and flight probably mediated these deterministic processes to influence the phenotypic structure of sympatric bats in urban landscapes.

Sociality confers many benefits to gregarious animals, but group-living does not come without associated costs. Social living can cause anxiety, which if prolonged can lead to stress. We investigate correlates of anxiety and stress among wild adult male olive baboons (Papio hamadryas anubis) in Gashaka-Gumti National Park, Nigeria, and present the first examination of these correlates for male baboons living in forested environments. We used rates of self-directed behaviours (SDB) to quantify anxiety, and faecal glucocorticoid (FGC) concentrations to quantify physiological stress. We first examined the relationship between SDB rates and FGC concentrations, and then examined each measure in relation to a range of variables previously linked to anxiety or stress. We found that SDB rates and FGC concentrations were not correlated with each other, nor were they related to any measure of grooming or receiving aggression. SDB rates were negatively correlated with time spent with an adult female as nearest neighbour, but were unrelated to time with another adult male as closest conspecific. FGC concentrations were unrelated to time with adult nearest neighbours of either sex. Our results add to the growing body of literature examining the psychological and physiological costs of social living among wild primates.

Morphological similarities between the tadpoles of Amietia umbraculata and A. vertebralis have led to confusion and incorrect descriptions and identifications in the literature. Based on 33 body measurements and ratios we revised the morphological descriptions of the tadpoles of the two species. Tadpole identification was verified through DNA sequencing using mitochondrial (16S) gene fragments. A combination of four morphological characters proved to be informative and consistent in distinguishing between tadpoles of the two species. Tadpoles of A. umbraculata are characterized by having four labial tooth rows in the lower jaw, extensive tail mottling, a dorsal fin that originates well behind the body, reaching a maximum depth at 50% of the tail length, and an average tail length of 1.9 times body length. Amietia vertebralis tadpoles on the other hand are characterized by having five or more labial tooth rows in the lower jaw, tail mottling that is confined to the upper half of the tail musculature, a dorsal fin that originates at the body-tail junction but retains a low profile, rising abruptly to reach a maximum depth at about 40% of the tail length, and an average tail length of 1.5 times body length. These four characters identify the two species without ambiguity.

The species richness of seven montane anuran communities of the Albertine Rift was assessed with visual and acoustic transects. A total of 28 anuran species were detected. The highland anuran communities at (2000–2800 m) were dominated by up to nine Albertine Rift endemic species considered threatened by IUCN Red List criteria, whereas the community at lower elevation included mainly wide-ranging, least-concern species. Cluster analyses of the presence/absence data on the overall array of species indicated that Ahl's reed frog (Hyperolius castaneus) is a suitable umbrella species for the highland Albertine Rift endemics. Hyperolius castaneus was always among the first species collected at each site, either by visual detection or by acoustic detection based on the characteristic advertisement and aggression calls which were emitted during the day and at night. The ease of its detection in the field makes this species an ideal candidate for rapid assessments of priority sites for conservation measures.

Ansell's mole-rat, Fukomys anselli, is a cooperatively breeding bathyergid endemic to the Lusaka Province of Zambia. During a 12-month field study involving the capture of 33 colonies of mole-rats, the number of occupants, breeding females and sex ratio within colonies were recorded. While thirty of these social groups contained a single breeding queen, three were found to exhibit plural breeding among females, having two queens present simultaneously. Mean ± S.E.M. colony size was 8.7 ± 2.2 (range 6–16, n = 33). The occurrence of plural breeding and small group size is consistent with the predictions of the aridity food distribution hypothesis for social species inhabiting mesic habitats. The sex ratio of the population was skewed towards females at 1.2:1. The mean mass of adult males was 63.0 ± 18.3 g (range 36.7–110.3, n = 87) while that for adult females was 52.9 ± 11.8 g (range 35.1–77.8, n = 86). Pregnant and lactating females were found throughout the study period from February 2009 to February 2010, indicating an aseasonal pattern of breeding in this species. Autopsy of individuals (n = 288) from the 33 colonies revealed a total of 19 pregnant females. Nine of these were in the latter stages of pregnancy and mean litter size was determined as 2.7 pups (range 1–4). Collectively, these new data reveal that Fukomys anselli share many similarities in life history and reproductive traits with the phylogentically closely related giant Zambian mole-rat, Fukomys mechowii.

The parasitic infections of amphibians in the National Park (pristine savanna), Buffer Zone (slightly altered savanna closer to fields) and Agricultural Zone (agricultural land including cotton fields with intensive use of pesticides) of the Pendjari Biosphere Reserve, Benin, were investigated. A total of 145 amphibian specimens of 14 species were examined and parasites recovered included two monogenean parasites, an acanthocephalan, two cestode species, four trematode species and eight nematode species. Results obtained show the possible influence of land-use pattern on parasite distribution. For example, the encysted acanthocephalan had higher prevalence in Phrynobatrachus latifrons caught in the Agricultural Zone, but specimens from the Buffer Zone had a higher intensity of infection. Infection with trematodes occurred predominantly in hosts collected in the Agricultural Zone, indicating that landscape alteration, pesticide use and nutrient enrichment from fertilizers in this zone may be enhancing intermediate host populations and hence parasite prevalence. Furthermore, this infection pattern may be indicative of an immunosuppressive effect of pesticides on the frogs of the Agricultural Zone. Prevalence of Rhabdias sp. infecting P. latifrons caught in the Buffer Zone was almost 14 times higher than those caught in the Agricultural Zone. Similarly, the infection with Amplicaecum sp. was higher in Amietophrynus maculatus from the National Park (57.1%) than those from the Agricultural Zone (15%). It is possible that the pesticide-contaminated environment of the Agricultural Zone inhibits the development of the free-living stages of these nematodes.

The yellow-bellied Scotophilus dinganii is the only African house bat species reported to occur in the Arabian Peninsula. Formerly, the Arabian house bats were referred to similar-looking white-bellied S. leucogaster, which differs from S. dinganii mainly by the colour of ventral pelage. We reassessed the taxonomic status of house bats from southwestern Yemen using genetic and morphological analyses. The Yemeni specimens clustered within two distantly related mitochondrial lineages of African Scotophilus: East African S. aff. dinganii, which is a paraphyletic group to S. dinganii s.str. from South Africa, and West African S. leucogaster. This taxonomic assignation was based on published sequences of reference museum specimens. Differences in external and cranial measurements also indicated the presence of two distinct taxa in Yemen. The Yemeni and comparative Ethiopian populations of S. aff. dinganii showed close morphological similarity to the type specimen of S. nigrita colias from Kenya. Because the Yemeni and Ethiopian yellow-bellied house bats cannot be synonymized with S. dinganii, the designation S. colias is tentatively suggested for this particular East African and Yemeni lineage of the S. dinganii complex. However, final correspondence of this name with the respective populations or applicability of some of other available names must yet be explored. Based on environmental differences of the Yemeni localities of origin, S. colias appears to be ecologically delimited to mountainous habitats, while S. leucogaster to harsh lowland deserts. This is consistent with known habitats of African populations of both species.

Arid and semi-arid environments are characterized by extreme fluctuations in temperature and low rainfall, which present significant challenges to the animals inhabiting these areas. However, the presence of burrows may allow animals to avoid climatic extremes and predators and may act as valuable foraging sites. We assessed the microhabitat conditions (maximum and minimum temperature, relative humidity and seed abundance) of aardvark (Orycteropus afer) burrows in relation to paired non-burrow areas at three sites in South Africa. We also describe the extent to which they are used as resources by other vertebrates. Maximum temperatures were significantly lower and minimum temperatures and relative humidity values were significantly higher inside the burrows than outside. The concentration of seeds inside the burrows and at the paired non-burrow sites was similar. Twenty-seven vertebrate species (21 mammals, two birds, three reptiles and one amphibian) were recorded making use of the burrows and it is likely that these species accrue benefits (e.g. a buffered microclimate) from burrow use. However, our sampling was biased towards mammals and nocturnal species. Consequently, we recommend further work to establish the overall reliance of vertebrate taxa on aardvark burrows in arid and semi-arid environments.

The parasitic gill monogenean Pseudodactylogyrus anguillae is alien to Africa. In an investigation of 227 longfin eel, Anguilla mossambica, and 26 mottled eel, Anguilla marmorata, sampled from four river systems in the Eastern Cape, South Africa, this parasite was only present on the gills of A. mossambica sampled from the Great Fish River system. In the Great Fish River, it infected 73.2% of the sampled population at a mean intensity of 63.8 ± 34.3 parasites per fish. High prevalence and intensity are indicative of a well-established alien invasive parasitic species. Results showing the absence of P. anguillae from the co-occurring but less abundant A. marmorata indicate that this species may be less susceptible to P. anguillae infection.

Largemouth bass (Micropterus salmoides) have been introduced to many South African river systems where they become invasive and pose a threat to native biota. The diets of small (32–138 mm TL) and large (192–448 mm TL) sized bass were analysed and compared in a marine fish nursery area in the lower Kowie River on the warm temperate coast of South Africa over a one-year period from March 2009 to February 2010. Dietary differences were detected between the two size groups. Amphipod sp. (% index of relative importance (IRI) = 69.2) and dipterans (Insecta) (%IRI = 21.9) dominated gut contents of small bass while larger bass preyed mostly on odonates (Insecta) (%IRI = 16.3) and the brachyuran Potamonautes sidneyi (%IRI = 80.0). Fish prey was of low importance during this study but comparisons with previous work on the lower Kowie River showed that when the river is flowing, young marine fish recruiting into the freshwater from the estuary become important prey items. These data suggest that in the lower Kowie River bass utilize invertebrate prey at low fish prey abundance and opportunistically feed on migrant fish when these are available.

Differences in the ecological niche requirements among rodent species competing in the same habitat may result from differences in the use of one to three resources: space, time and food or some combination of these. Alternatively, differences in resource use utilization among animal species may simply reflect availability of food, and when food is limited, different animal species compete. In this study, the diet of two rodent pest species, Mastomys natalensis and Gerbilliscus vicinus, coexisting in fallow land in central Tanzania were studied to assess the degree of diet differentiation among them. Dietary niche breadth of G. vicinus was greater than that of M. natalensis in all stages of the maize cropping seasons. The rodent species studied overlapped considerably in the food items consumed ranging from niche overlap (O_jk) of 0.77–0.89. Grains/seeds featured high in the diet of M. natalensis while plant material occurrence was high in G. vicinus. These two food categories may have contributed to differences in diet partitioning, which may, in turn, facilitate their coexistence in fallow land.

As part of a three-year study to develop ecologically-based rodent management (EBRM) in southern Africa, a capture—mark—recapture study was carried out in Tanzania, Namibia and Swaziland to establish the demographic patterns and population dynamics of rodents. Two study grids were established in each country. In Tanzania, ten species of rodents and one shrew (Crocidura sp.) were identified in the study area. The rodent species consisted of Mastomys natalensis, Aethomys chrysophilus, Arvicanthis neumanni, Gerbilliscus vicina, Acomys spinosissimus, Lemniscomys griselda, Lemniscomys zebra, Rattus rattus, Graphiurus sp. and Mus minutoides. Mastomys natalensis was dominant and contributed more than 70% of the captures. In Namibia, five species were captured, namely Mastomys natalensis, Gerbilliscus leucogaster, Saccostomus campestris, Mus minutoides and Steatomys pratensis. Mastomys natalensis contributed about 50% of the captures. In Swaziland, only M. natalensis was captured in the study grids. There was a clear pattern in the population dynamics, with breeding confined to the wet seasons in the three countries. Mastomys natalensis was the dominant pest species, for which EBRM should focus on. The highest population density of M. natalensis occurred during and after the rains, which coincided with the most susceptible phenological stage of crops. The breeding seasonality and density fluctuations observed in the three countries conform with observations made elsewhere in Africa, which support the hypothesis that rainfall events promote primary productivity on which murid rodents depend. Development of EBRM in these countries will be determined by the local conditions and how they influence the demographic processes of the rodent populations. EBRM should make use of the available ecological knowledge of the local rodent pest species and the focus should be on (ecological) management practices applicable at the community level including community based intensive trapping, field hygiene, removing cover and sources of food for rodents.

Germinated and non-germinated cereals (paddy, Oryza sativa; pearl millet, Pennisetum typhoides; and ragi, Eleusine coracana) and pulses (green gram, Phaseolus aureus; black gram, Phaseolus mungo; and Bengal gram, Cicer arietinum) were evaluated to determine their suitability as a bait base for zinc phosphide and bromadiolone for controlling Rattus rattus. Two-choice and multiple-choice experiments were conducted using 10 animals of mixed sex for each experiment to compare bait preference of germinated and non-germinated grains. The results of the two-choice tests revealed that both germinated cereals and pulses were preferred to non-germinated by the test animals. The differences between the quantities consumed were statistically significant (P < 0.01). The results of Duncan's post hoc multiple comparison tests of germinated grain consumption by R. rattus were categorized into three subsets. The first subset consisted of pearl millet, Bengal gram, ragi and paddy. The second and third subsets consisted of green gram and black gram, respectively. It is concluded that by considering the cost of all four grains (in the first subset) any one of the cheapest among them in the germinated form could be recommended or used as a bait base for any of the two rodenticides for the management of R. rattus. Based on these results, germinated paddy was chosen (as it is the cheapest among the grains) for the laboratory toxicity studies. The acute toxicity tests (2% zinc phosphide) resulted in 100% mortality of rats under both no-choice and two-choice (germinated paddy with poison versus plain non-germinated paddy bait) tests. A powder formulation of bromadiolone (0.005%)-treated bait yielded 100% mortality of rats in no-choice tests, but only 60% mortality in two-choice tests. These results suggest that the 2% zinc phosphide with germinated paddy as a bait base was preferred over a powder formulation of bromadiolone (0.005%). However, a field efficacy study is required before a final bait formulation recommendation is made.

Two species of echiurans (Echiura), Anelassorhynchus inanensis (Ikeda, 1904) and Ochetostoma pellucidum (Fischer, 1895), were collected from intertidal rocky shores at Perriers Rock on the KwaZulu-Natal north coast. This is the first known record of these two species from South African waters and considerably extends their geographical ranges. A. inanensis, originally described by Ikeda (1904) from Naha, Japan, was later recorded and redescribed by Wesenberg-Lund (1939) from a single specimen from South Annam while Haldar & DattaGupta (1991) assigned two specimens from Kavaratti Island, on the southwest coast of India, to the species. O. pellucidum is based on several specimens collected from Whydah, West Africa. According to Stephen & Edmonds (1972) no illustrations of this species have been given while the description given by Wesenberg-Lund (1954) is based on a single badly-preserved specimen from French Guinea. Both species are redescribed and compared with other closely related forms.

In the present study the genetic variation of the Critically Endangered velvet worm species Opisthopatus roseus is examined. This species is endemic to the Ngele mistbelt forest in the KwaZulu-Natal province of South Africa. In recent years the forest has been severely impacted by anthropogenic activities such as logging of indigenous trees, construction of a national highway, commercial timber plantations and introduction of alien plant species, resulting in habitat fragmentation and potential range contraction for the species. A total of 35 specimens were collected from five sample sites (logs) in the Ngele Forest and sequenced for a partial fragment of the cytochrome c oxidase subunit one (COI) locus. In addition, the sex ratio at each sample sites was determined. A highly skewed sex ratio was detected within sites and only three males were present among the 35 O. roseus specimens. The results reveal the presence of 11 haplotypes and moderate genetic differentiation between the five sampled sites. In addition, it is evident that samples from the same log represented a mixture of haplotypes and did not comprise a single maternal haplotype. The results further suggest that gene flow between individuals occurs in the interior of the forest. Conservation implications of the present study are briefly discussed.

Despite widespread and increasing use of solar-powered satellite transmitters to tag wild birds, there are few published articles that detail how transmitters should be attached to different species and even fewer assessments of the overall field success of telemetry projects. The scarcity of this information makes it difficult to plan and budget for telemetry projects effectively. In this paper we present relevant information from a study involving a total of 47 individual ducks of two Afrotropical anatid species, Egyptian Goose, Alopochen aegyptiaca, and red-billed teal, Anas erythrorhyncha, using solarpowered GPS satellite transmitters of two different sizes (30 g and 22 g, respectively) at three very different southern African sites (Strandfontein wastewater treatment works in the Western Cape Province of South Africa, Barberspan Nature Reserve in the North West Province of South Africa, and Lake Manyame in north-central Zimbabwe). We present a full description of harness design and attachment and a survivorship analysis of the transmitters. Our results suggest that the 30 g units last longer than the 22 g units, with approximately 60% and 30%, respectively, of these PTTs (position tracking terminals) lasting longer than a year; 45% and 5%, respectively, lasting longer than two years; and 20% and 0%, respectively, lasting longer than three years. We strongly encourage the publication of comparable data sets so that future studies that rely on telemetry data can be planned with a realistic set of assumptions and limitations in mind.