The migratory flight behaviour of Nilaparvata lugens (Stål), other delphacid and cicadellid pests of rice and some of their heteropteran predators was investigated during the dry-season crop in an irrigated rice-growing area of the Philippines. A combination of radar (including an X-band and a newly-developed Q-band system (8 mm wavelength)), aerial netting, ground trapping and visual observations was used to determine diel changes in aerial density and composition of arthropods in flight over the study site. The most abundant migrant caught in the aerial nets was Cyrtorhinus lividipennis Reuter (a mirid predator of N. lugens), with Nephotettix spp. and Nilaparvata lugens the next most abundant species. Flight activity in all migrants was found to be very largely confined to periods of about 30 min at dusk and dawn, with minimal activity at other times of the day and night. Migration distance in the dusk flight was generally limited to 6–30 km, depending on wind speed, and the dawn migration covered shorter distances due to the lighter winds. Small numbers of rice cicadellids and delphacids including N. lugens were occasionally detected later in the night, and these may have had the potential to migrate longer distances.

Survival rates of females of Musca vetustissima Walker were estimated as a function of their reproductive age, based on age-distributions of field females whose headwidths ranged from 1·5 to 2·5 mm. Females had a mean life expectancy of 140 day-degrees above 8°C, which was independent of fly size. They did not respond to baited traps during the first 15 day-degrees C of adult life. Thereafter, trappability varied with reproductive age but was affected by fly size only during the initial stage of the first ovarian cycle. Because ovarian development rates and female fecundity are dependent on fly size, the expected numbers of eggs laid by field females 1·5–2·5 mm in headwidth during their lifetime range from 28 to 106.

Investigations into feeding behaviour and activity patterns, basically by biconical trap sampling, were carried out as part of a study of the population ecology and dynamics of Glossina pallidipes Austen in the Lambwe Valley, Kenya. Host-preference determinations from blood-meal analyses showed that over 80% of feeds constantly derived from bushbuck, buffalo and bushpig, in roughly equal proportions. Activity cycles in relation to feeding and pregnancy, investigated by mark-release-recapture and physiological methods, indicated four-day feeding intervals for males and three-day ones for females. The periodicity was more variable in males, and in females was contingent upon reproductive events, with feeding mainly taking place just before the in-utero moult of the second-instar larva to the third instar, just after larviposition and again at a more variable time during pregnancy. Comparisons of nutritional status from size, weight, and fat and haematin levels showed that flies from thicket had consistently greater fat reserves than flies from Acacia woodland and conifer plantation. No seasonal differences were found, nor any relationships between other variables. Based on hourly sample data, both sexes showed unimodal diel activity patterns, but with clear differences. Activity was most closely correlated with temperature, to a lesser extent with saturation deficit and least with light intensity, and better correlations were obtained with females than males. Diel activity patterns in both sexes were also age-related, and from the close similarity found between male and virgin female activity patterns it was inferred that male activity around traps was partly sexually oriented. Using electric screens in conjunction with biconical traps, comparisons between flies attracted to traps and those actually caught showed that traps were relatively more efficient in catching males than females, but captures were representative of the active population in terms of age and pregnancy.

Leafhoppers of the genus Cicadulina transmit pathogens to Gramineae (including cereals and sugarcane) in Africa, India and South America. Previous taxonomic and applied studies on Cicadulina are reviewed, and an identification key to the 22 recognizable species is given; each species is discussed and illustrated. Two species, C. dabrowskii sp. n. and C. immaculata sp. n., are described; C. anomala Van Rensburg, C. kalaharica Van Rensburg, C. inopia Van Rensburg, C. ungula Van Rensburg, C. bispinosa Van Rensburg, C. opaca Van Rensburg, C. waltersi Van Rensburg, C. scabra Van Rensburg and C. letabae Van Rensburg are established as junior synonyms of C. anestae Van Rensburg, C. bipunctella (Matsumura) as a junior synonym of C. bipunctata (Melichar) and C. triangula Ruppel as a junior synonym of C. storeyi China.

A number of tettigoniids belonging to the genera Sexava, Segestes and Segestidea are responsible for severe defoliation of coconut palm, oil palm and Pandanus spp. in Papua New Guinea. Most chemical and cultural methods of control are either impractical or uneconomic. The possibility of biological control was examined with reference to some of the natural enemies of Segestes decoratus Redtenbacher, in particular a parasite of nymphs and adults, Stichotrema dallatorreanum Hofeneder, as well as the egg parasites Triteleia atrella (Dodd), a species of Tetrastichus near T. dubius (Waterston) and a second species of Tetrastichus. It was concluded that S. dallatorreanum contributed to the control of Segestes decoratus in areas with an evenly distributed rainfall and that the parasite had potential as a biological control agent in areas where it is not endemic against pest species of Sexava, Segestes and Segestidea. A survey of the egg parasites showed that they were not effective in controlling the pests as they were only able to parasitize eggs in exposed oviposition sites. Egg predators of unknown identity were responsible for higher egg mortality than were the egg parasites.

Multivariate techniques, principally the method of canonical variates, were used to investigate morphological variation within and between populations of the group of Myzus persicae (Sulzer) from North and South America, Europe, Africa and Asia. The scores on the first canonical variate of samples from tobacco in North America, the Mediterranean region, the Middle East, Africa and Sri Lanka all grouped consistently when compared with samples from other host-plants, even after aphids from tobacco had been reared for up to seven years on a non-tobacco host. Thus there is a widely-distributed tobacco-adapted form, closely related to M. persicae but with its own characteristic morphology. Morphological discriminants are given for the recognition of apterous and alate viviparae of this tobacco form, which is given the name M. nicotianae sp. n. Both M. persicae and M. nicotianae have 2n = 12, and both are frequently heterozygous for apparently the same autosomal translocation, which they must have acquired independently. M. nicotianae is presumably isolated from M. persicae by being permanently parthenogenetic. In Japan and Central Asia, however, aphids of the M. persicae group on tobacco can produce sexual morphs; the taxonomic status of these latter populations is still unclear. Multivariate comparison of European and North American populations of dark green aphids of the M. persicae group with 13 or 14 chromosomes in somatic cell nuclei instead of the normal 12, led to the conclusion that these all belong to one morphologically variable taxon, M. antirrhinii (Macchiati). Keys are provided to the apterous and alate virginoparae of the species of the M. persicae group in America.

The identity of Phenacoccus gossypii Townsend & Cockerell is established as a local species in Mexico and southern USA. Although P. madeirensis Green was described from Madeira and is recorded from much of the African mainland, it was probably introduced from tropical America. It is now known to be common on many crop and ornamental plants in North and South America where it has often been mistaken for P. gossypii. Both species are redescribed and illustrated, and their range of variation is discussed. P. grenadensis Green & Laing and P. harbisoni Peterson are synonymized with P. madeirensis. A related species, P. baccharidis sp. n., is described from Brazil, and P. sphaeralceae sp. n. is described from Arizona. P. hurdi McKenzie is redescribed and is now recorded from an extensive area in southern USA and Mexico. All five species have numerous dorsal multilocular pores and are included in a key to 15 related species in North and South America.

The mortality of eggs and larvae of Musca vetustissima Walker, caused by two scarabaeine dung beetles was measured in the laboratory in favourable cattle dung at 23°C. Eggs and larvae less than one day old were the main stages killed in the dung. Mortality from beetles that entered dung immediately after fly oviposition was higher than that from beetles that entered the dung one or three days later. The results also indicated that there are differences between species of dung beetles in the mode of action against the fly. The smaller beetle species, Onthophagus binodis (Thunberg), at densities of 120 beetles/pad caused increased mortalities of 12% of fly eggs and 11% of young larvae but did not affect older larvae, while the larger species, O. ferox Harold, at densities of 48 beetles/pad caused no extra mortality of eggs but did cause an increased mortality of 47% of young fly larvae and 61% of older larvae.

A method is presented for quantifying moth area densities in individual crops and average densities over the total crop area. Mortality within the total Heliothis armigera (Hübner) adult population as a result of insecticide applications against larvae on cotton in the Sudan Gezira was found to be very substantial for males and virgin and inseminated females. The level of adult mortality which was previously unknown, was shown to be an important factor in Heliothis control. The relative importance of alternative host crops as source sites of adults which infest cotton was demonstrated. Initially, larval populations built up in groundnut fields between July and early September. Females emerging from groundnut fields colonized sorghum in late August and early September and cotton slightly later. Larval densities found in groundnut and sorghum fields correlated strongly with the average densities of virgin females found 12 days later. Groundnuts provided the most important source of adults. The virgin female population was in turn highly correlated with the remainder of the adult population (males and inseminated females), demonstrating that H. armigera adults attacking cotton within the Gezira are locally produced. Adults were attracted to sorghum in large numbers for feeding and oviposition, but survival there of older larvae (fourth to sixth instars) to the adult stage was only 25% of that in groundnuts. The implications for pest management are discussed.

Field and cage observations were made on Heliothis armigera (Hübner) in the Sudan Gezira to quantify nocturnal patterns of moth emergence, copulation and oviposition and the ages at which copulation and egg-laying occur. The flight activity pattern of males, virgins and inseminated females was assessed for crops with scarce or abundant sources of honeydew or nectar. A pronounced redistribution of moths between crops during the peak flight activity period (18.15-19.30 h) was demonstrated and measured, with the net exchange being approximately balanced. The direction of this movement was dependent upon the attractiveness of crops for feeding or oviposition. Male and female moths developing from larvae reared on groundnuts had significantly shorter lifespans when fed only water and less fat-body at emergence than moths developing from larvae reared on sorghum. When offered sugar solution, lifespans were significantly greater and differences in the lifespan of moths from groundnuts or sorghum were no longer evident. Adult feeding appears to compensate for any deficiencies associated with a larval diet of groundnut. Feeding was important for adult survival; there was a strong attraction of moths to crops with plentiful honeydew or nectar. A descriptive behavioural model is presented for the first half of adult life, until peak oviposition, and the uses of this model for pest management are discussed.