An island of 4·5 km2 in Lake Kariba, Zimbabwe, was stocked with cattle and infested with Glossina morsitans morsitans Westwood and G. pallidipes Austen in 1979. From February 1980 to April 1981, the tsetse populations, estimated by mark, release and recapture, increased about ten times, to contain about 3000 males of G. m. morsitans and 2000 males of G. pallidipes. From May 1981 to May 1983, six traps, with carbon dioxide and acetone as odour attractants, were used to capture 0·1–0·3% per day of the G. m. morsitans population and 1–4% per day of the G. pallidipes population. Captured flies were retained and killed, or were automatically sterilized with metepa and released. In May 1983, when the populations of G. m. morsitans and G. pallidipes had declined by about 90 and 99%, respectively, the traps were replaced by 20 targets with acetone and 1-octen-3-ol as attractants. The targets were coated with dieldrin or, later, deltamethrin and killed about 2% per day of G. m. morsitans and 5% per day of G. pallidipes. Both populations then declined rapidly, G. pallidipes disappearing in 11 weeks and G. m. morsitans in nine months. Targets offer a cheap, simple and effective means of eliminating isolated populations of tsetse.

Chilozela trapeziana (Sepp) comb. n. is the senior name for the pyralid whose larvae damage cassava leaves in South America, and Botys jatrophalis Walker syn. n. and Phlyctaenodes bifilalis Hampson syn. n. are junior synonyms.

Field studies of Mythimna convecta (Walker), Persectania ewingii (Westwood) and P. dyscrita Common, and their associated parasitoids, were conducted in cereals and pastures in Victoria, Australia, in 1980–83. The study included a survey of the armyworm fauna in the major agricultural districts, detailed phenological studies within five oat crops, and ultraviolet-light trapping of adults. M. convecta was the most abundant species in cereals and was found more commonly in the warmer months of spring and summer in all districts. Larval development in oat crops during spring was preceded by little moth activity in light traps. P. ewingii was common in the southern districts and in the cooler months from May to September. It was the most prevalent armyworm species in pastures and rarely occurred in barley. P. dyscrita was the least abundant species and occurred mostly in the drier inland areas, particularly in the west. Light-trap catches of P. ewingii and P. dyscrita suggested there were two generations, one in spring and one in the autumn. Regular sampling in oat crops showed that M. convecta developed faster and was usually more abundant than P. ewingii during seed development and ripening, and hence had the potential to inflict more crop damage. Of the 17 species of parasitoids reared from field-collected larvae, the wasp Campoletis sp. was the most abundant. In 1980–82, average parasitism ranged from 13 to 16% but fell to 2% after the drought of 1982.

Five capture-recapture experiments with three species of the group of Anopheles punctulatus Dönitz were undertaken in four villages in Madang Province, Papua New Guinea. Females were collected in landing catches, blood-fed, marked with fluorescent powder according to time and location of capture, and released. Recaptures took place for 4–13 nights after release. A proportion of the recaptured mosquitoes was dissected for evidence of recent oviposition and gonotrophic age determination. Two models were developed to test a number of the assumptions that underlie the analysis of capture-recapture data from mosquitoes. Species abundance and numbers collected varied within and between villages, but marked mosquitoes dispersed at random within the mosquito population. No heterogeneity was found between indoor and outdoor biters of any of the three species. Survival rates were similar in all species. Significant differences were found in the duration of the oviposition cycle between species and village. This was 2·7–3·7 days for A. punctulatus, 2·4–3·2 days for A. koliensis Owen and 2·1–3·0 days for A. farauti Laveran. No differences were found in the duration of the oviposition cycle of nulliparous and parous females of A. farauti. Thus temporary pool breeders had a longer cycle than permanent pool breeders. This tended to be due to a delay in returning to feed rather than a delay in oviposition. It is concluded that differences between local environments exert a greater influence on the duration of the oviposition cycle than do species specific differences.

Ticks were regularly collected from cattle along transects in Sukumaland and the Southern Highlands, Tanzania, and from locations near Tabora, Mbeya, Arusha and Dar es Salaam from 1973 to 1976. Marked seasonal variation in abundance occurred in Rhipicephalus appendiculatus Neumann in the Southern Highlands (but not in Sukumaland) and in Amblyomma variegatum (F.) near Tabora. It was possible to detect Theileria parva antibodies in cattle sera from the Southern Highlands only during the season of R. appendiculatus adult abundance. Despite this there was no evidence of enzootic instability of the disease. The results demonstrate that the behaviour and distributions of these and the other species of ticks found are not fixed and constant but vary according to a complicated interplay of factors as yet imperfectly understood, such as climate and vegetation and host density, susceptibility and grazing habits.

There is a growing body of evidence to suggest that low density populations of Spodoptera exempta (Walker) in the solitary-phase can persist during the dry season in certain localities in eastern Africa. Moths were caught in simple pheromone traps for several months after the estimated dates of emergence of larval outbreaks in central Tanzania in 1980 and from 1982 to 1984. This provided evidence for the occurrence of a shifting population between the Uluguru Mountains and the swampy coastal areas, in east-central Tanzania, for a large part of the year. Central Tanzania is an important area for the development of the first outbreaks of the season in eastern Africa and for any control strategy. It has been suggested that S. exempta fits the criteria for an r-selected species. However, the low migratory capacity of low-density populations indicates rather that the species spreads out its risks in space and time. A theoretical genetic mechanism controlling the distance flown by moths may enable the species to exhibit both migratory and non-migratory strategies. Thus both ephemeral and continously favourable habitats may be exploited.

Parasitoids of Aonidiella aurantii (Maskell) on citrus in South Africa were monitored using two types of yellow sticky trap. One of these traps was highly efficient, being fluorescent with peak reflectance at about 530 nm. Aphytis spp. populations were low before February and high thereafter. Citrus surrounded by natural bush was an isolated reservoir of high host and parasitoid population levels. Aphytis spatial distribution within the orchard was extremely patchy, with over 100-fold differences in population levels over a distance of a few metres. This patchiness mirrored that of its host. This contagious spatial pattern was maintained despite 1000-fold seasonal changes in population levels. These temporal changes were characteristic and general throughout an orchard, and independent of patchiness. Initial Aphytis population levels did not dictate the final population level at the end of the season. Comperiella bifasciata Howard and its hyperparasitoid Marietta javensis (Howard) also showed clear seasonal population trends, but not of the same magnitude as those of Aphytis. There was no statistically significant correlation between the spatial distribution of one parasitoid with that of another, even between C. bifasciata and M. javensis. The patchiness of these two species was not correlated with overall host density. Aphytis and C. bifasciata were partially mutually exclusive. Aphytis was by far the most economically important of the parasitoids. Pest management practices, therefore, should aim at conserving the pool of Aphytis within the orchard as far as practicable.

The integrated use of biconical traps, insecticide-impregnated targets and the sterile insect technique was developed for the eradication of Glossina palpalis palpalis (Robineau-Desvoidy) in a 1500-km2 area of central Nigeria. Six weeks or more of continuous removal trapping, using biconical traps, reduced the target tsetse population by more than 90% but failed to eradicate it. Males sterilized by irradiation from a 60Co source that were then released weekly induced significant sterility in target females and were successful in helping to eradicate the target population. A minimum ratio of 10:1 of sterile to wild males was required to achieve eradication in a central area of 300 km2. In marginal habitats, insecticide-impregnated targets were found adequate to control the tsetse population. The targets were also efficient as barriers to prevent reinvasion of the area. The combined effect of removal trapping and sterile male release is expected to eradicate G. p. palpalis from the entire study area.

Sitophilus oryzae (L.) is a serious pest of stored cereal grains, but recently a few populations have been found able to feed and breed on certain legumes also. Legumes contain many toxic phytochemicals. A technique is described to generate data suitable for probit analysis, enabling a dose- response line to a toxic food to be constructed. The toxic effects of two legumes, yellow split-pea and adzuki bean, on adult survival and reproduction in three geographical strains of S. oryzae were examined. The strains were shown to vary in their abilities to tolerate legume phytotoxins, but there was no correlation between the abilities to survive on different legumes in the same strain. Yellow split-pea was shown to have residual effects on adult survival and reproduction, but performance was enhanced rather than impaired. It is suggested that improved performance results from induction of certain enzymes by substances within yellow split-pea. The use of legumes mixed with cereals as a control measure against S. oryzae is also discussed.

An inexpensive and versatile method for maintaining age-structured Musca domestica L. populations for studies on the evolution of insecticide resistance is described. Adult flies are kept in spacious aluminium cages in which age-structuring is maintained by the thrice-weekly addition of pupae bred from eggs collected from within the cages. The population size is regulated in a density-independent manner by constraining the input of pupae to that necessary to maintain the required equilibrium density of adults within a cage. Adult numbers are monitored by photographing from outside the cage flies settled on a grid etched on the rear wall, and by converting this grid count to an estimate of population size using a calculated regression line. Observed changes in fly numbers in a cage accorded well with those predicted by a computer model using empirical data on larval productivity, and the emergence and survivorship schedules of adult flies. Since a variety of insecticide control regimes can be applied within the cage, this system enables an adequate yet tractable simulation of selection for resistance by insecticides under field conditions.

Anopheles minimus Theobald was studied in four villages that had experienced differing levels of DDT application during malaria control operations in the foothills of northern Thailand. During times of high population level, the behaviour of the mosquito was assessed with respect to man-biting indoors and outdoors and feeding from domestic bovids. Its responses in villages recently treated with DDT differed from those in villages that had not been treated for several years, so that the insecticide had little direct lethal effect. Furthermore, the influence of the insecticide on the behaviour of the insect was considerably longer than is the lethal effect of the chemical. There was a suggestion that the taxon A. minimus may include two or more morphologically-cryptic species.

Females of Anopheles minimus Theobald at a site in northern Thailand that were caught on man or domestic bovids and were released showed a significant tendency to return to the type of host upon which they were first caught (P = 0·03). From a total of 3526 individuals marked and released, 51 were recaptured. A simple explanation is that the taxon is in fact a mixture of two or more morphologically cryptic species, a suggestion already indicated by previous studies.

A sandwich enzyme-linked immunosorbent assay (ELISA) test was developed to detect blood-meals in insects and identify the host fed on. The test proved both sensitive and specific. Very small quantities of fresh blood (about 0·02 μl) can be detected; in practice, this enables blood in mosquitoes which are about three-quarters gravid to be identified. In trials in both Zambia and Britain, positive reactions were easily identified visually; consequently, this enabled the ELISA technique to be used as a routine field test. In addition to those of mosquitoes, blood-meals of a few Culicoides species were also successfully identified.

An attempt to eradicate Glossina pallidipes Austen from the Lambwe Valley of western Kenya by sequential aerial application of endosulfan aerosol spray is described. The operation was ultimately unsuccessful, even after nine sprays of relatively heavy dosages of insecticide. Survivors were present throughout after each spray, and females mated and bred between sprays. The population was reduced by over 99·9% in main habitats of thicket and woodland and by about 90% in conifer plantation. The outcome of a model of spray effectiveness suggested, however, that the final reduction in population size was the cumulative effect of only about a 90% kill per spray application. Post-spray monitoring showed that the tsetse population in thicket returned to its pre-spray level in little more than 12 months and became stabilized thereafter. Elsewhere, recovery was much slower and more variable. Evidence was found for population regulation by density-dependent emigration of flies. Calculations of population growth rate gave an instantaneous rate of increase of 0·0148/day, equivalent to a finite rate of increase of 1·568/month or a doubling time of 47 days. This permitted a 90-fold increase in population size over a ten-month period of unrestricted growth. The failure to achieve eradication is discussed in terms of certain operational shortcomings and environmental constraints which exist in the Lambwe Valley.

Since Maconellicoccus hirsutus (Green), a pest of many plants including Hibiscus, was introduced into Egypt in 1908, probably from the Oriental Region, it has spread to much of tropical Africa. There is little information available on its control, but the encyrtid Anagyrus kamali Moursi is known to give good control in Egypt. The ladybird Cryptolaemus montrouzieri Mulsant is a voracious feeder on M. hirsutus in Egypt, but it does not survive the Egyptian winter, although it may be useful in the tropical areas. M. perforatus (De Lotto) is synonymized with M. hirsutus, and the distribution and host-plants of this species in Africa are discussed. A lectotype is designated for M. ugandae (Laing), a species now known to occur throughout tropical Africa, especially on coffee and cocoa. It is illustrated and redescribed and compared with M. hirsutus. M. bardus De Lotto, known from South Africa only, is excluded from the study because it probably belongs to the genus Chorizococcus.