Observations were made on crowded, laboratory-reared. 18–20-day-old males of Schistocerca gregaria (Forsk.) taking off in a wind tunnel. Depending upon the time of day, take-off, which was usually upwind, particularly at higher wind speeds, was related to food availability. Few locusts with full foreguts took off, especially at low air temperatures (20°C), while at high temperatures (35°C), 95% of ones starved for 6 h took off. At 30°C, there was little further increase with starvation for 24–30 h. Females containing large oocytes took off less readily than those with small ones. Take-off was also sometimes initiated by decreases in wind speed and light intensity, and sudden changes in air temperature. A distinct passing shadow stimulated take-off, but not the flickering shadows of other flying locusts. A period of flight immediately beforehand inhibited take-off. The results are discussed in relation to field observations.

To improve the precision of understanding of the phenology of Listronotus bonariensis (Kuschel), samples were collected at regular intervals near Canterbury, New Zealand, and the seasonal changes in their reproductive morphology monitored by regular dissection. It was demonstrated that the adult weevils enter hibernatory reproductive diapause in early March and remain in this state until late July when reproductive activity resumes. Adults emerging in February may enter diapause immediately. The existence of diapause was further supported by its associated cessation of mating, increase in body fat and apparent drop in the haemolymph titre of juvenile hormone. Laboratory experiments indicated that a critical photoperiod induces diapause and an eventual reduction in sensitivity to photoperiod allows reproductive activity to resume.

The insect fauna of a block of Amelonado cocoa trees at Tafo, Ghana, partly shaded by Gliricidia sepium and partly unshaded was sampled. Crematogaster clariventris (Mayr) and Oecopyhlla longinoda (Latr.) were the most numerous species of ants, the former occurring on 84% of the trees and the latter on 17%. There was a considerable overlap in the distribution of the two species. Thirty-two other species of ant were collected by pyrethrum knockdown sprays from the cocoa canopy, and 47 species were collected from leaf litter beneath the trees. More species of ant were recovered from litter on shaded plots than on unshaded ones. There was some avoidance of trees occupied by O. longinoda by certain canopy .ants. A greater number of Lepidoptera larvae, Orthoptera and the mirid Sahlbergella singularis Hagl were collected by pyrethrum knockdown sprays from unshaded plots than from shaded ones, whereas Diptera and Hymenoptera other than ants were more numerous on shaded plots. The most abundant coccid was Stictococcus sjostedti Ckll., a species tended for its honeydew by both dominant ants; it was more common on shaded plots. Mealybug species, of which Planococcoides njalensis (Laing) was the most common, were more abundant on unshaded plots.

Seasonal variation in the circadian periodicity of mating behaviour in Chilo suppressalis (Wlk.) was investigated. In the field, mating was initiated at dusk in June, but not until after dark in August. The times of females calling and of male catches in a pheromone trap were later in August than in June. The seasonal variation in time of male catches coincided approximately with that of mating, suggesting that latter may be attributable in part to seasonal changes in the response of males to the sex pheromone. Also, it is likely that these variations are affected by environmental factors, especially temperature.

The diurnal flight and biting activity patterns of Culicoides phlebotomus (Will.) were investigated during ten all-day (06·00–19·00 h) trapping sessions at Las Cuevas Bay, north Trinidad. A comparison was made of the results obtained by biting catches, paddle trapping (in which a piece of greased wire screen was swung in front of the collector's legs) and truck trapping. Biting activity persisted throughout the day, with peaks of activity one to two hours after sunrise and again at sunset. Truck trapping recorded little variation in flight activity of empty females during daylight hours, but that of engorged, half-gravid and gravid females was greater during the afternoon than in the morning. Male flight activity was high in the early morning and mid-afternoon but declined rapidly towards sunset. Paddle trapping alone recorded a substantial increase in the blood-seeking activity of parous females after sunset. The paddle-trapping technique was found to be a useful alternative to biting collections for the study of anthropophilic Culicoides.

Spodoptera exempta (Wlk.) was reared for one year at three locations in Kenya: Msabaha on the coast, Nairobi, and Mbita Point on Lake Victoria. The potential yearly number of generations was 15·8 at Msabaha, 6·6 at Nairobi and 13·1 at Mbita Point. Average developmental time from egg deposition to adult emergence was 23·1, 54·9 and 27·8 days, respectively. The longest, 77·3 days, was recorded in Nairobi and the shortest, 19·4 days, at Msabaha. The survival rate varied little seasonally at the coast but was erratic inland, with some generations failing and others very successful. Larvae offered a choice of maize and stargrass (Cynodon dactylon) showed a preference for stargrass. Survival was higher on stargrass than on maize but the developmental time was longer. Mortality was highest in the larval stage. In all generations and at all locations, death was usually caused by a nuclear polyhedrosis virus. There was a positive relationship between larval mortality and rainfall, and a negative one between larval mortality and sunshine. Mating was restricted to after midnight. Females oviposited and then mated during the same night. Peak egg-laying took place before midnight, with a smaller peak at about 02·00–04·00 h.

Aedes ledgeri Huang, A. soleatus Edw. and A. calceatus Edw. of the group of A. aegypti (L.) of the subgenus Stegomyia collected on the coast of Kenya are distinguished on the strength of ecological and genetic (isozyme) data. The preferred oviposition sites of A. ledgeri and A. soleatus were tree holes above ground level, whereas A. calceatus favoured abandoned gastropod shells. A. ledgeri was most abundant in ecotonal habitats and A. soleatus in undisturbed forest At least two diagnostic enzyme loci separated each of the three species pairs.

The effect of insecticide drift on Glossina morsitans centralis Machado in Botswana was assessed using two-man fly-rounds downwind of the 1700–4000 km2 aerial spraying blocks. The blocks were subjected to sequential applications of endosulfan at between 6 and 12 g/ha with a droplet v.m.d. of 30 μm. Post-spray reinvasion of flies was assessed using two-man fly-round surveys across the edge of one spray block along ideal linear habitat. Each application had an effect on the flies outside the spray block, even at a distance of 20 km downwind. The first two applications of a series had a greater effect than later ones. Total reductions of G. m. centralis after a series of 5 or 6 applications were 92·5% 7 km outside a block sprayed in 1975, 98·3% 6–12 km outside a block sprayed in 1977 and 64·2% 10–20 km outside a block sprayed in 1978. Ageing studies showed that some females 10–20 km downwind of the sprayed area survived up to five applications, and three months after the final application the population there had recovered almost to pre-spray levels. The main front of fly encroachment after spraying across the downwind edge of the 1978 block travelled at about 0·8 km/month in the first 17 months after spraying, but small numbers of flies were found well ahead of the main front. It was concluded that the situation in the drift area could be ideal for the development of insecticide resistance and that adjacent blocks sprayed in consecutive years should overlap by 10 km.

A soft scale found infesting mango in Israel is described from the female as Coccus elatensis sp. n. It is morphologically related to C. gymnospori (Green), which is redescribed.

A population of Aedes aegypti (L.) near the Kenya coast was monitored for nearly a year after a genetic control experiment. Two genetic markers that were carried by released males but unknown to the region persisted in the population during the entire period of observation. The endurance of introduced genetic material in a natural population is an important step toward the use of genetic control in insect management.

The toxicities at 25°C and 70% RH of methyl bromide at about 4 mg/litre and phosphine at about 0·24 mg/litre were determined for large populations, including all developmental stages, of a number of strains of seven species of stored product beetles from 29 countries for a range of exposure periods. The species tested were Oryzaephilus surmamensis (L.), Rhyzopertha dominica (F.), Sitophilus granarius (L.), S. oryzae (L.), S. zeamais Motsch., Tribolium castaneum (Hbst.) and T. confusum Duv. With both fumigants, some individuals of certain strains of S. oryzae and T. castaneum survived a concentration × time product twice that sufficient to kill every individual of other strains of these species, and this was also true for R. dominica with methyl bromide and T. confusum with phosphine. With methyl bromide, some individuals of six species survived the longest exposure used. For the two species of Tribolium, this was 48 h, and for the three of Sitophilus and R. dominica it was 32 h. With phosphine, individuals in many strains of S. granarius, S. oryzae and S. zeamais survived for 8 days. In several of the species, there was some correlation between the tolerances of immature stages and of adults for both fumigants. Many strains were tolerant to both fumigants.

Megaselia sandhui sp. n. is described from specimens of the adult male, female and pupa infesting cultivated mushrooms in India. It is closely related to the other species of Megaselia known to be pests of cultivated mushrooms.

The sterilising effect of 90Sr β-radiation was tested against males and females of Glossina palpalis palpalis (R.-D.) 1–6 days old. The flies were exposed to doses ranging from 2 to 40 krad at a dose-rate of l·09 × 103 rad/min. The optimal sterilising dose for the males was about 7·5 krad. Females that mated with males that had been exposed to this dose failed to produce larvae although the sperm of the treated males was motile. There was no relation between dose and length of life at doses from 2 to 10 krad but at higher doses survival was greatly reduced. Females survived irradiation better than males.

The behaviour of Myzus persicae (Sulz.) on plant surfaces treated with emulsions of saturated fatty acids was investigated in three types of test. Aphids showed no clear preferences for settling or larviposition when the acids were applied to whole Chinese cabbage plants or to leaf discs from several different plant species, but settling on halves of Chinese cabbage leaves treated with undecanoic, dodecanoic or heptadecanoic acids was significantly decreased.

A laboratory evaluation of growth, food consumption and food utilisation by larvae of Spodoptera littoralis (Boisd.) was made from the beginning of the fifth instar to the prepupal stage on cabbage, cotton, kenaf and Congo jute (Urena lobata). Growth, measured by daily larval fresh weight, was highest on cabbage and similar on kenaf, Urena and cotton. Development was fastest on cabbage and slowest on cotton and Urena. Larvae underwent an extra moult on the latter two food-plants. Weight gain was highest on cabbage, followed by cotton, Urena and kenaf, in that order. Total food consumption was highest on cotton and lowest on kenaf and Urena. Cabbage was most efficiently digested, followed by kenaf, cotton and Urena, in that order. Digestive efficiency on all four plants declined with larval age. The efficiency of conversion of ingested food to body material was highest on cabbage and lowest on cotton and kenaf. Digested portions of Urena were most efficiently converted to body substance while digested portions of cabbage and kenaf were most inefficiently converted to body material. It is concluded that S. littoralis is better adapted to cabbage than any of the other three plants and that increasing cultivation of cabbage in Ghana is likely to enhance the pest status of the insect.

In studies in southern England, the relative abundance of the hymenopterous parasites attacking Metopolophium dirhodum (Wlk.) and Sitobion avenae (F.) differed between their hosts, between years (1973–79) and on whether they were assessed using field or laboratory-reared mummies. Species of the complex of Aphidius uzbekistanicus Luzhetzki were usually the commonest parasites, especially from mummies of M. dirhodum, while Aphidius picipes (Nees) was less common and tended to attack mainly S. avenae; Toxares deltiger (Hal.) was common only in 1979. Species of hyperparasites emerged most frequently from mummies of M. dirhodum, except for Dendrocerus carpenteri (Curt.), which emerged most often from S. avenae.

Epigaeic ants were sampled by pitfall trapping in three South African citrus orchards supporting low levels of Aonidiella aurantii(Mask.). All three were similar except that chemical control of the scale was in progress in one, biological control in the second, while the third was in transition from chemical to biological control. Within-orchard variation was statistically zero at the chemical and biological control sites, both of which had constant species abundances and rank-abundance sequences throughout. Within the transition site, rank-abundance sequences were constant but species abundancies varied. Despite the statistical constancy of rank, within-orchard samples at the transition and chemical control sites varied in their dominant ants. The biological control site was dominated throughout by Pheidole megacephala(F.). Individuals of the dominant species made up between 59 and 69% of the total, and as the dominant species varied, this high figure suggested strong interspecific competition. Log-normal models were fitted to the site rank-abundance plots. The interquartile index, Q, was higher when based on an underlying log-series model (α) than on a log-normal one; it was about 2·0, which is low compared with most other studies. The total numbers of individuals and species of ants were higher in the biological control than chemical control orchard. Diversity, as measured by α or Q log-normalwas lowest in the chemical control site, medium in the biological controlone, and highest in the transition site. Between-site variation showedvariation in general ant abundance. As with the within-orchard results, those between orchards were relatively constant. Ant management in relation to community structure is critically appraised on the basis of limiting factors (especially abundance of honeydew as food for potentially dominant ants) and rank-abundance sequences.

The distribution and seasonal activity of Onthophagus granulatus Boh. was studied at two sites in eastern Australia. It is a univoltine species, with peaks of activity in spring and early summer. Breeding cycles and periods of stress were demonstrated by the proportions of newly emerged, nulliparous and parous beetles and those resorbing oocytes. The optimum temperature for brood production was 25°C. The threshold of development was 11·3°C, and 495 day-degrees C were required for development from egg to adult. Dung quality and drought were important factors affecting survival and brood production in the field. The distribution of the species in south-eastern Australia is limited by summer rainfall and temperature.

Differences in population growth rates, r, and net reproductive rates, R0, are shown between Glossina palpalis gambiensis Vanderplank, G. palpalis palpalis (R.-D.), G. morsitans morsitans Westw., G. morsitans centralis Machado, G. pallidipes Aust. and G. fuscipes fuscipes Newst. from field-collected data and laboratory colonies. The relationships of these population parameters with temperature are shown as well as the temperature-dependence of T, the cohort generation time. Two rapid methods for estimating r from R0 knowing the mean temperature, and a more accurate rapid method of calculating r direct from the capacity for increase, re, were developed from these data. The stable age-structure, C, calculated from these data was used to show that the laboratory-derived age-specific fecundity, mx, is 22% higher than in the field. The difference is probably due to natural abortion in the field and the better nutritional status of laboratory-reared flies. Deviation from the stable age-structure is shown to be an indication of the sample bias and population equilibrium. The relationships of these parameters to the state of a given population and to optimal timing of control efforts are considered.

Goniozus antipodum Westw. is redescribed, and three species, G. jacintae sp. n., G. collessi sp. n. and G. mandibulatus sp. n., are described from the females, figured and keyed. All are from Australia, and G. jacintae and G. mandibulatus are parasites of the tortricid Epiphyas postvittana Wlk.