Callosobruchus subinnotatus (Pic) is the major insect pest of stored bambara groundnuts, Vigna subterranea (L.) Verdcourt, in sub-Saharan West Africa, but little is currently known about its biology or how it may be controlled. A series of laboratory studies was performed to investigate the bionomics of and differences between two apparently different morphs of adult of each sex of this species, here termed ‘active’ and ‘normal’. Major differences in their morphology, physiology and behaviour were identified and are described in detail for the first time. They provide clear evidence of the existence of an adult polymorphism among populations of this species, which is comparable in certain respects to that previously described for C. maculatus (Fabricius) and C. chinensis Linnaeus. Adults can be separated into the correct morph based on characteristic differences in elytral and pygidial colour and pattern. ‘Normal’ adults are characterized by having high fecundity, short adult life and are relatively sedentary while ‘active’ adults exhibit reproductive diapause (suspension of reproductive activity), are long lived, and show (at least in females) increased dispersal tendencies. These characteristics suggest adaptation of the ‘active’and ‘normal’ morphs respectively to the different environments of field and seed stores, and the significance of the polymorphism in the life history of C. subinnotatus is discussed in this context. The design of any effective control regime for this bruchid needs to take account of and could potentially be based upon the existence of polymorphism in C. subinnotatus.

A laboratory study of aspects of parasitoid host acceptance, suitability and physiological regulation in natural and novel host species was carried out to investigate the degree of variability encountered with different hosts and to determine the value of such observations in host range determination. The parasitoid Microctonus aethiopoides Loan was exposed to a natural host, Sitona discoideus Gyllenhal (Coleoptera: Curculionidae) and three novel hosts, the New Zealand native Nicaeana cervinaBroun, the introduced weed biological control agent Rhinocyllus conicus (Froehlich), and a congeneric pest species, Sitona lepidus Gyllenhal (all Coleoptera: Curculionidae). Per cent parasitism of these species was 54%, 43%, 39% and 0%, respectively. The results indicated that for both S. discoideus and R. conicus more males than females were parasitized (69% cf. 45%, and 49% cf. 32% respectively) but host size was not a significant factor. Overall, superparasitism was recorded in about 29% of parasitized weevils and there was evidence that host discrimination to avoid superparasitism occurred in the natural host. Conversely, superparasitism occurred more frequently than would be expected in N. cervina (42%) coupled with higher survival of larvae in superparasitized hosts in this species. The frequency distribution of attack of R. conicus by M. aethiopoides was not different from random. Parasitoid development was more rapid in the natural host, S. discoideus, and parasitoid size was positively correlated with host size. There was a strong positive relationship between parasitoid larval survival and the presence of teratocytes in all hosts. Host fecundity and fertility were reduced by parasitism for most species, and in some cases, by exposure to parasitoids in the absence of detectable parasitism. It was concluded that laboratory observations can provide useful information on the compatibility between host and parasitoid which can complement traditional host range tests to predict field host range.

Various aspects were studied of the brood size and sex allocation strategies, and of size-fitness relationships in Parallorhogas pyralophagus (Marsh), a gregarious ectoparasitoid of Eoreuma loftini Dyar. Brood size was significantly correlated with host size; larger hosts were allocated larger broods. Brood sex ratios were fixed precisely at 1 male per 4 females, and eggs were likely to be deposited in that order; differential mortality did not contribute to this precise sex ratio. The sex allocation strategy of P. pyralophagus is likely to conform to strict, i.e. single foundress, local mate competition. Adoption of this strategy is probably influenced by a limited insemination capacity of males; a smaller proportion of females (0.09 vs. 0.21) remained virgin in broods with precise or higher sex ratios (≥0.20 males) relative to broods with lower than precise sex ratios (< 0.20 males). Moreover, all females were inseminated in most broods (60%) with precise or higher sex ratios, whereas this did not occur in broods with lower than precise sex ratios. The hypothesized occurrence of strict local mate competition in P. pyralophagus was supported also by observations that: (i) offspring brood sex ratios were independent of maternal brood sex ratios and number of parental females concurrently allocating offspring to a group of hosts, and; (ii) the rate of superparasitism under no-choice conditions was low (~20%), suggesting that rates of outbreeding in the field are low. Other results suggested that fitness in P. pyralophaguswas correlated with adult size; longevity and reproductive capacity both increased with adult size in males and females. However, adult size may be more important for females than for males because the differences in reproductive capacity between the largest and smallest individuals was up to 7.3 times greater in females versus <2 times in males.

Anopheles funestus Giles has been implicated as a major malaria vector in sub-Saharan Africa where pyrethroid insecticides are widely used in agriculture and public health. Samples of this species from northern Kwazulu/Natal in South Africa and the Beluluane region of southern Mozambique showed evidence of resistance to pyrethroid insecticides. Insecticide exposure, synergist and biochemical assays conducted on A. funestus suggested that elevated levels of mixed function oxidases were responsible for the detoxification of pyrethroids in resistant mosquitoes in these areas. The data suggested that this mechanism was also conferring cross-resistance to the carbamate insecticide propoxur.

The suitability of ten diets for the development and reproduction of Franklinothrips orizabensis Johansen, the key natural enemy of Scirtothrips perseae Nakahara, a pest of California grown avocados, was determined in the laboratory. The experimental diets evaluated were: (i) irradiated Ephestia kuehniella Zeller eggs; (ii) irradiated E. kuehniella eggs and avocado pollen; (iii) Tetranychus pacificusMcGregor eggs; (iv) T. pacificus eggs and avocado pollen; (v) irradiated E. kuehniella eggs and T. pacificus eggs; (vi) irradiated E. kuehniellaeggs, T. pacificus eggs and avocado pollen; (vii) Scirtothrips perseae; (viii) Heliothrips haemorrhoidalis (Bouchè); (ix) avocado pollen; and (x) a young avocado leaf. Franklinothrips orizabensis larvae were unable to develop to adulthood on diets 9 and 10. The remaining eight diets supported complete development of F. orizabensis, but only diets 1, 2, 5, 6, 7 and 8 produced fecund females. On diet 5, F. orizabensis exhibited high larval to adult survivorship (90%), mated females exhibited highest daily and lifetime fecundity, and the progeny of mated females were female biased (53%). Analysis of jackknife estimates of net reproduction (Ro), intrinsic rate of increase (rm), and finite rate of increase (λ) were all significantly greater for F. orizabensisreared on irradiated E. kuehniella eggs and T. pacificus eggs (i.e. diet 5) than corresponding values for other diets on which female F. orizabensiswere able to complete development and reproduce. Incorporation of avocado pollen into diets had an adverse effect on demographic statistics for F. orizabensis, and low quality diets resulted in male biased sex ratios for this predator.

To monitor the quality of male tsetse for use in the sterile insect technique (SIT), a field cage test was developed and evaluated. Mating competitiveness was tested with male Glossina pallidipes Austen that emerged from pupae stored for different periods at 15°C. Control males emerged from pupae stored at 23–24°C and emerged at 26.5°C. Each sample of test males was divided into two groups with one group being irradiated at 120 Gy; the other group was not irradiated. More than 70% of the maximum possible number of mating pairs occurred in all tests. Males emerged from pupae kept at low temperature and then irradiated formed a greater proportion of mating pairs than the controls. Males emerged from pupae kept at 15°C generally started mating more quickly than the standard colony males although there was no significant difference. Insemination rates were above 99%. Pooled data indicated that mean spermathecal values for females mated with irradiated males were significantly lower than for control males. The duration of copulation varied significantly between treatment groups and was significantly longer for irradiated male flies; there was no correlation between duration of copulation and mean spermathecal value.

A dual port olfactometer was used to study the response of Anopheles gambiaeGiles sensu stricto to odours of human and animal origin. Human odour consisted of human skin emanations collected on a nylon stocking, which was worn for 24 h. This was tested alone or together with 4.5% carbon dioxide, the concentration in human and cattle breath. Cattle odours consisted of cow skin emanations and/or carbon dioxide. Cow skin emanations were collected by tying a nylon stocking (‘cow sock’) around the hind leg of a cow for 12 h. Anopheles gambiae s.s. was consistently highly attracted by human odour, which is consistent with the high degree of anthropophily in this mosquito. Anopheles gambiae s.s. was not attracted by human or cattle equivalent volumes of carbon dioxide and this gas did not enhance the effect of human skin residues. Furthermore, A. gambiae s.s. showed a high degree of aversion to cow odour. When human odour and cow odour were tested together in the same port, mosquitoes were still highly attracted, indicating that whilst cattle odour may deter A. gambiae s.s., these mosquitoes can detect human odour in the presence of cattle odour. It was concluded that carbon dioxide plays a minor role in the host seeking behaviour of A. gambiae s.s., whilst host specific cues such as human skin residues play a major role and very effectively demonstrated anthropophilic behaviour in the laboratory.

The two ‘forms’ of the Afrotropical shoot-fly Atherigona rubricornis Stein are shown to be good species. A neotype is designated for the true A. rubricornis, which is restricted to tropical Africa. The other form is described as A. triticisp. n. It is widespread in tropical Africa, but is known also from Egypt, where the larvae attack wheat seedlings, Israel and Yemen. Infestation levels of larvae, with maxima between 9% and 32%, indicate that it is a significant pest of wheat seedlings in Egypt.

The effects of two natural aphid enemies, adult Coccinella septempunctataLinneaus, a predator, and Aphidius rhopalosiphi de Stefani Perez, a parasitoid, on spread of barley yellow dwarf virus (BYDV) transmitted by the bird cherry–oat aphid, Rhopalosiphum padi (Linnaeus) were studied under laboratory conditions. Predators or parasitoids were introduced to trays of durum wheat seedlings and the patterns of virus infection were observed after two, seven and 14 days of exposure. More plants were infected with BYDV in control trays without A. rhopalosiphithan in trays with the parasitoid present, both seven and 14 days after the introduction of parasitoids. Patterns of virus infection were found to be similar over time in trays with a parasitoid present and in control trays. More plants were infected in trays with C. septempunctata present than in control trays, both two and seven days after the introduction of the coccinellid. The spread of virus infections progressed differently over time for the two treatments (predator and parasitoid), differences between treatments being most marked after two days and seven days, when more plants exposed to predators but fewer exposed to parasitoids were infected with BYDV compared to their respective controls. However, by the 14th day 88% of all plants were infected and there was no significant difference between the two treatments. The role of natural enemies in spread of BYDV is discussed.

The population dynamics of the white wax scale, Ceroplastes destructor Newstead, was studied intensively in four easy-peel citrus orchards in the Western Cape Province of South Africa over three consecutive years (1997–1999). Key factor analysis was used to determine and quantify the contribution of individual mortality factors to the total generation mortality. Key stage mortality, determined from a cohort life table, was in the third instar and pre-ovipositing female stages. Mortality of C. destructor was caused primarily by parasitoids, predators and miscellaneous factors. Parasitoids and miscellaneous factors acted as density-dependent regulatory agents during the pre-ovipositional and first instar stages respectively. This has implications for biological control of C. destructor. Some of the mortality factors acted either randomly with no reference to the population densities or in an inverse density-dependent manner during the egg-crawler, second or third instar stages.