Dispersal flights of termites are the result of the sum of a number of complex and highly orchestrated behavioural responses by different castes to a variety of cues. Macrotermes natalensis (Haviland) produced a single brood of alates each year. These alates left their parent nests in dispersal flights early in the summer rainy season from late September to early December, with a peak during October. Rainfall was the primary stimulatory factor triggering flight; 5 mm was the threshold for initial flights. Large flights usually occurred later in the season and required more rain to trigger them. The initiation and duration of individual flights were regulated by factors such as temperature, light intensity, and wind. The threshold temperature for swarming flights lay between 17–19°C. If the air temperature dropped below this the flight was aborted. The species did not fly during rain and major flights took place between 39 % and 90 % relative humidity. Windless conditions were favoured for flight which ensured that alates were not distributed so widely that the chances of finding a mate were jeopardized and that the female sex pheromone was not so dispersed that males failed to locate mates. Flight times varied over the flight season as they were governed by light intensity and took place at dusk in low light intensities with alates leaving the nest from 1 to 30 lux. The majority of swarming flights occurred with a waxing moon.

Foraging behaviour and the diurnal and seasonal foraging periodicity of the southern African snouted harvester termite, Trinervitermes trinervoides (Sjöstedt), are described. The termite usually foraged during summer nights when ambient temperatures were 10–14 °C. Foraging occurred at temperatures between 13–25 °C under laboratory conditions. A foraging party emerging from a single hole harvested over an area of approximately 0.78 m². A colony exploited a total area of approximately 214 m² and foraging areas of adjacent colonies overlapped. Colonies occupy single mounds.

We analysed the distribution of living larval Chironomidae (Insecta: Diptera) along a depth transect (0–80 m water depth) at Kigoma Bay in Lake Tanganyika (East Africa) to explore the ecological indicator value of Lake Tanganyika's midge fauna and to delineate the habitat preferences of resident larvae. Additionally, by comparing the taxonomic resolution reached in live versus fossil chironomid studies in this lake, we appraised the general information content of African fossil chironomid assemblages and, therefore, the quality of African chironomid-based palaeoenvironmental reconstructions. We found 141 chironomid specimens, representing two Tanypodinae, one Orthocladinae and 13 Chironominae (seven Chironomini and six Tanytarsini) taxa. All of these are conspecific to larval types encountered in subfossil collections from Lake Tanganyika or smaller East African lakes, but their differential diagnoses are updated here to include exoskeletal elements that are generally better preserved on live specimens. The chironomid assemblage collected in the Kigoma Bay depth samples is dominated by a few species of chironomine detritus and algae feeders, concentrated in areas of high macrophyte density. Additionally, there is a small component of tanypodine predatory taxa, along with several relatively rare, possibly habitat-specialist, taxa of both subfamilies. Our results further indicate that littoral and sublittoral palaeoenvironments in Lake Tanganyika may generally be inferred from high fossil density and diversity. The fairly high correspondence between inferred habitat preference of fossil and live collections in Lake Tanganyika indicates that, in large African lakes, the near-to-offshore gradient in chironomid habitat is imprinted on fossil assemblages. Hence, death assemblages deposited recently at various locations and microhabitats within Lake Tanganyika can be used as modern calibration data for palaeodepth inference. Moreover, Tanganyikan fossil-based taxonomy fully matches the taxonomic resolution reached in this study, so ecological information retrieved from live specimens can immediately be used for palaeoenvironmental studies.

Used individually, pesticides, biological control agents, cultural practices and host plant resistance have either not been effective or are not sustainable in control of the sorghum shoot fly, Atherigona soccata Rondani (Diptera: Muscidae). An experiment, with seed sown in September and October 2002, to determine the effect of a combination of variety, fertilizer, insecticide and time of planting on incidence of A. soccata, and yield, and yield components of sorghum, Sorghum bicolor (L.) Moench, was conducted in Ugunja, western Kenya. Two varieties of sorghum were used, Ohunjo and Serena, susceptible and resistant varieties, respectively. Although results were variable amongst the different combination of factors tested, it was found that incidence of A. soccata was lowered and sorghum grain yield increased through the manipulation of management practices such as genotype selection, fertilizer and insecticide application.

Common bean, Phaseolus vulgaris L., is the most widely grown legume in Kenya. The black bean aphid, Aphis fabae Scopoli, is a sporadic but serious pest of beans, causing significant yield losses in Kenya mainly during the short rainy season. Growing resistant and tolerant bean varieties is the most sustainable method for controlling the aphid, but information on varietal resistance is lacking. Studies were conducted on the varieties grown by Kenyan farmers to identify which are resistant or tolerant to the aphid during the short rainy seasons of 2002/03 and 2004/05. Ten varieties were evaluated in a randomized complete block design. Dimethoate (insecticide) kept the control, which comprised the entire ten varieties, aphid-free. Bean variety significantly affected aphid infestation in 2002/03 but not in 2004/05. Significant yield reduction as a result of aphid infestation was observed in five varieties in 2002/03 and all but one in 2004/05. Correct selection of varieties to grow during the short rainy season could therefore lead to high yields of common beans.

Patterns of damage by Cryptophlebia peltastica (Meyrick) (Lepidoptera: Tortricidae) on litchi fruits, Litchi chinensis Sonnerat, at varying stages of maturity were determined in four orchards in Mauritius during two consecutive litchi seasons. Damage was found to start at a very early stage in the season, when fruits were 3 mm in diameter. In three orchards there were no significant differences in damage incidence by C. peltastica during the fruiting season. In the fourth orchard, a significant increase in fruit damage was observed at the end of the season. Damage levels of up to 20 % on litchi fruits were observed when no control actions were implemented.

Four tortricid species are of major economic importance on deciduous fruit tree crops in South Africa. They are codling moth, Cydia pomonella, oriental fruit moth, Grapholita molesta, false codling moth, Thaumatotibia leucotreta, and carnation worm, Epichoristodes acerbella. Pest management and phytosanitary practices are often hampered by the inability to distinguish between the immature stages of these four species. To address this need, the final instar larvae and pupae of E. acerbella are described and illustrated. The morphology of these immature stages of E. acerbella is compared with those of the aforementioned tortricid species. Keys are included to distinguish between the four species. In addition, as an alternative method of identification of any life stage, a means of discriminating between the four species, based on analysis of nucleotide sequences of the mitochondrial cytochrome oxidase I gene, is described.

Toxicity of the essential oil from the leaves of Mentha longifolia (Lamiaceae) was evaluated against Sitophilus zeamais, an insect that is the major cause of deterioration in stored grains, including maize. The contact, fumigation and repellency bioassays revealed that the oil was toxic to the insect. Increases in oil concentration and time of exposure resulted in a progressive increase in insect mortality. The dose of essential oil at 0.50 μl/g of maize grains in the contact bioassay caused 100 % mortality of S. zeamais, compared to less than 10 % mortality recorded at 0.125 μl/g. The essential oil demonstrated moderate fumigation toxicity against the test insects at concentrations of 24 and 32 μl of oil per litre of air, respectively. A high repellency was recorded for the oil at all the concentrations tested. Repellency values of as high as 100 % were recorded for most concentrations. Mentha longifolia could be a potential agent for the protection of agricultural stored products against S. zeamais.

The morphology of the puparia of Curtonotum saheliense Tsacas and C. simile Tsacas are described or redescribed, based on material reared from locust egg pods in Mali and Oman, using SEM techniques and light microscopy. The original specimens on which Greathead based his 1958 larval description of C. cuthbertsoni Duda, have been located and examined and are here ascribed to C. simile, rather than C. saheliense as previously assumed, based on examination of the type specimens of both. An additional larva, also reared from locust egg pods in northern Nigeria, is tentatively ascribed to C. saheliense and is described using SEM techniques, despite its relatively poor state of preservation. The two species are shown to be separable based on the relative lengths and widths of the mouthooks and intermediate sclerite of the cephaloskeleton and in the degree of expansion of semi-circular cuticular folding adjacent to the anal opening on the anal pad. A diagnosis of the immature stages of the genus Curtonotum Macquart is provided and the larval and puparial features of C. saheliense and C. simile are compared to the puparium of the Nearctic species C. helvum (Loew), the other species of the genus for which the immature stages are partially known. The composition and homologies of caudal tubules are discussed, as are features in common between the three species under discussion, and tentative comparison is made of the known immature stages of Curtonotum to those of the known larvae of other families of the Ephydroidea, as additional larval features are now discernible for the Curtonotidae. It is concluded that the larval morphology of the Curtonotidae is generally close to the putative groundplan of the Ephydroidea and are probably most closely allied to the Drosophilidae and Camillidae. The relative length of abdominal compartments six and seven is probably an autapomorphy of the family.

Variability in the spermathecal form of the three closely-related species ascribed to the Curtonotum cuthbertsoni species-group, cuthbertsoni complex, namely: C. cuthbertsoni Duda, C. saheliense Tsacas and C. simile Tsacas, is investigated. Pairs of spermathecae are illustrated photographically and their form compared intra- and interspecifically. The external form and shape are found to overlap in all three species. The spermathecae of C. simile are illustrated for the first time. While general trends in shape are noted, variation is overlapping, both between individuals of the same species and within pairs of spermathecae from the same specimen. It is concluded, therefore, that spermathecal form is not a useful character to distinguish females of the three species, and that the use of spermathecae as diagnostic characters in the genus Curtonotum Macquart, in general, must be viewed critically and with a high degree of scepticism.

Conservation of South Africa's butterflies is critically dependent on knowledge of their ecology. Many endangered species are in the family Lycaenidae and have a more-or-less obligate relationship with a particular species of ant. An ecosystem capable of hosting such butterfly species must also be able to sustain substantial populations of the host ant. The Brenton blue, Orachrysops niobe, is one of the most endangered South African butterflies, and is known only from one site in the Western Cape. Ant assemblages were investigated at this site and two other ecologically similar sites where the host plant of O. niobe, Indigofera erecta, also occurs, but O. niobe was absent. The ant assemblages differed significantly, with the host ant of O. niobe, Camponotus baynei, only occurring at the Brenton site where the butterfly was breeding. Vegetation composition, structure and microclimate differed at the three sites and these appear to be key factors in determining the ant assemblages that a site will support. The two sites where C. baynei was absent had a history of recent fire, which contributed to the vegetation changes and deprived the ant of dead wood for nesting. The invasive Argentine ant, Linepithema humile, was not detected at any of the study sites, and the reasons for this are discussed.

This study was undertaken to determine whether image texture derived from high spatial resolution imagery could successfully predict Sirex noctilio infestation levels in a pine plantation forest in the eastern part of South Africa. The woodwasp, S. noctilio, damages trees by releasing toxic mucus into the wood during oviposition. A 50 × 50 m grid was generated over the study area and 65 cells were randomly selected. Within each cell, a 10 m circular plot was created and the level (%) of S. noctilio infestation was visually assessed. Using high spatial resolution imagery (0.5 m), 13 texture measures were calculated with various window sizes. Correlation tests were then performed to determine the relationship between image texture and S. noctilio infestation levels. Individual correlations were average in strength. Consequently, a stepwise multiple linear regression analysis was undertaken to determine whether a selection of texture images could result in a more accurate prediction of S. noctilio infestation levels. A strong correlation of r = 0.7 between the predicted and the observed levels of S. noctilio infestation using the developed multiple linear regression model showed that image texture is a promising tool for the detection, and ultimately mapping, of S. noctilio infestation in plantations. The result is critical for improving the management of insect infestation in plantations of South Africa.

The status of the 15 species of Asterolecaniidae previously described or recorded from South Africa is reviewed, and the adult female and immature stages of a new species, Asterolecanium endroedyi, are described. Several species are transferred to other genera, representing new combinations, i.e. Asterodiaspis borboniae (Brain), A. brevispinum (Brain), Russellaspis brachylenae (Brain), R. conspicuum (Brain), R. euryopis (Fuller), R. euphorbiae (Russell) and R. proteae (Giliomee & Munting). Lectotypes have been designated for A. brevispinum, Planchonia stentae (Brain), R. brachylenae, R. conspicuum, and R. euryopis. Keys to the adult females of the genera and species of Asterolecaniidae known from South Africa based on the adult female are provided.

The larvae of Anchophthalmus algoensis Péringuey, 1904 and A. silphoides Gerstaecker, 1854 are described and illustrated for the first time, based on preserved material in the Transvaal Museum (National Flagship Institution) in Pretoria (33 larvae). The great similarity of the morphological structures of the larvae of A. silphoides and A. algoensis made it impossible to distinguish them. A key to larvae of known subtribe Platynotina is presented.