The effects of wind speed, wind direction, strong moonlight and rainfall on hourly light-trap catches of adult Spodoptera exempta (Wlk.) made at Muguga, Kenya, between January and June in 1970 and 1971 were investigated. The highest catches occurred in winds of 2–3 m/s, but generally, fewer moths were caught as wind speed increased. In totally dark hours the geometric mean catch at wind speeds of 5–6 m/s was about one-tenth that at 0–1 m/s. The relationship between catch and wind speed changed between January and June. Wind direction and change in direction were not significantly related to catch. The geometric means of catches made in strong moonlight were about one-tenth those in total darkness at comparable wind speeds. Rainfall was associated with higher catches in seven out of eight months, but the differences were significant in only two months. Recognition of the association between light winds and large numbers of S. exempta should assist forecasting, but a fuller understanding of the relationship between light-trap catch and population requires a method of population assessment that is independent of the effects of weather and moonlight.

The eleven Melanesian species of the genera Perisphincter Townes and Agrypon Foerster, endophagous parasites of Lepidoptera larvae, are revised, and a key to them is given. One species of Perisphincter, P. toxeres sp. n., and six species of Agrypon, A. augrhantum sp. n., A. ambomum sp. n., A. celenellum sp. n., A. rhamphidum sp. n., A. apistum sp. n. and A. phlogeum sp. n., are described. A. nigricans Szépligeti is placed as a synonym of A. caliginosum (Tosquinet).

Field collections in Bahia State, Brazil, of open cocoa flowers from an untreated plantation and one treated with insecticide and fungicide showed that Homoptera, including the cicadellid Xesrocephalus ancorifer Linnavuori, the aphidid Toxoptera aurantii (Boy.), and species of Pseudococcidae and Margarodidae (Coccoidea), constituted 96% of the total number of insects collected. In the untreated area, Wasmannia rochai Ford formed 22% of the ants collected and Brachymyrmex pictus Mayr, Crernatogaster ? parabiotica Forel,Iridomyrmex sp., Pheidole sp. and Solenopsis sp., a further 48%, whereas in the treated area, 47% were C. parabiotica, 14% W. rochai and 9% B. heeri Forel. There was evidence of a mosaic distribution pattern for the ant species. No cocoa pollen was found on any of the insects collected. Total pollination ranged between 7·4 and 8·1 % and effective pollination between 3·7 and 2·8 % in the treated and untreated areas, respectively. Approximately 76% of the pollen masses were stylar, with 38% of the ball type and 54% of the smear type, whereas the 16% stigmatic pollen masses consisted of 63% ball and 29% smear types; 44% of the stigmatic pollinations were effective, compared with only 33 % of the stylar ones. Some quadrats were more attractive to pollinators than others, and there was a significant positive correlation between percentage effective pollination and insolation, probably due to the favourable effect of sunshine on Forcipomyia activity.

Zonocerus variegatus (L.) is a chronic but relatively minor pest of crops in West Africa, and expensive control measures are not generally warranted. Egg pods, which take approximately six months to develop, are deposited in highly aggregated sites found scattered throughout cultivated areas. To test the feasibility of reducing populations in large areas by digging up the egg pods, two areas, of 18·5 and 20 ha, were selected near Ibadan, Nigeria, for investigation. Egg-laying sites were identified by observation and the egg pods subsequently dug up. There were on average one or two sites, 5·7 m2 in area, per hectare. Counting the following season's hatching populations in the two areas showed that the potential hatches had been reduced by 83 and 91%, respectively. Control by community self-help is suggested, where individual farmers identify and dig up laying sites on their own land, exposing the egg pods to desiccation on the soil surface. If this was carried out by all the farmers in an area, populations of Zonocerus could be reduced and crop damage minimised. Such control can be carried out at little or no cost and with very little labour.

An investigation using modified Manitoba traps on a forest track in Scotland showed that carbon dioxide, coloured spheres, movement, heat, Haematopota pluvialis (L.) and a swarm of other adults of Hydrotaea irritans (Fall.) were attractive. Traps with coloured spheres attracted more flies than traps without spheres. Matt black, shiny black and shiny red spheres attracted more males and females than did shiny green, white or yellow spheres. Significant increases in the number of flies caught did not occur when carbon dioxide was released from a trap at rates above 2 litres/min. Carbon dioxide seems to act as a long-range attractant, with vision becoming important near the host. Manitoba traps were also used with three repellents for H. irritans in order to evaluate the trap's potential for testing repellents.

Individuals and groups of adults of Zonocerus variegates (L.) were fed on various food-plants. Somatic growth, maturation and reproduction were all faster in group-fed insects than in isolated ones. Dry-season insects grew faster and produced greater numbers of egg pods than did wet-season ones. Short-winged insects developed faster than long-winged ones. Zonocerus that fed on cassava (Manihot esculenra) had a shorter somatic growth period and produced greater numbers of egg pods than insects that fed on any other foods. Cassava-fed Zonocerus had a greater efficiency of conversion of digested food to body substance, particularly during oocyte growth, than similar insects given other foods. Food plants such as Citrus and A spilia were inefficiently utilised, and often compensatory increases in consumption occurred. The success of the large dryseason population of Zonocerus in southern Nigeria is probably due to the increase in cassava cultivation.

Flies attracted to odours from 500 to 11 500 kg of livestock were captured using electrified netting in the Zambesi Valley, Rhodesia. Catches of maleGlossina morsitansWestw., Tabanidae, Stomoxyinae and non-biting muscids reached maximum levels at odour concentrations produced by only 500–3500 kg of livestock, but catches of femaleG. morsitans and both sexes of G. pallidipes Aust. increased over the whole range of concentrations tested. With the maximum dose of odour, up to 7100 tsetse were caught in 3 h. The proportion of tenerals in catches of Glossina was not affected by odour concentration. Daily catches of Giossina were positively correlated with daily maximum temperatures. The maintenance of large catches of flies for 60 days and the indication that larger catches are possible suggest the feasibility of employing odour-baited traps for tsetse control.

When incorporated into an artificial diet or painted on the surface of the membrane containing the diet, solutions of the insect repellent GD 880 (2-methyl-2-(octylamino)-1-propanol) killed aphids, whilst extracts of neem seeds (Azadirachta indica) or extracts of the bodies ofMyzus persicae (Sulz.) deterred settling and larviposition of apterous adults. The deterrent effect occurred even when the painted membrane was covered by a similar but untreated membrane. When applied in culture solutions to the roots of young kale plants, GD 880 and extracts of neem influenced aphids on the foliage only at concentrations that visibly affected plant growth, but neem extract (and to a lesser extent GD 880) decreased aphid colonisation when painted on kale leaves at non-damaging concentrations.

Felled logs of twelve species of Eucalyptus were exposed simultaneously to attack by Phoracantha semipunctata (F.) in Malawi. Colonisation was assessed by counting the number of larval feeding channels on the surface of the sapwood. The level of colonisation varied significantly between different tree species. E. fastigata and E. paniculata were much more heavily colonised than any of the other species. Possible factors contributing to this variation in susceptibility are discussed. The more susceptible species should be the most efficient for control operations involving the use of trap trees.

On Mahé in the Seychelles, Anoplolepis longipes (Jerd.) is a pest, mainly because of the nuisance it causes to man and domestic animals but also because it increases the numbers of sap-feeding insects and the occurrence of sooty mould on plants. The most important sap insect associated with sooty mould on citrus and cinnamon was Ceroplastes rubens Mask., the numbers of which sometimes increased a hundredfold in the presence of A. longipes, and up to 90% of leaves were infected with sooty mould. The rapid population expansion of the ant on Mahé prior to 1973 has since slowed down, and infestations in some areas have diminished, possibly because of a decreasing source of protein. Movement of ant-infested material has initiated secondary infestations, which are difficult to eradicate once established. The ant is not considered a threat to conservation as differences between the fauna in infested and uninfested areas tend to be quantitative rather than qualitative. Attempts to eradicate the ant on Mahé would be costly and inappropriate. The use of toxic baits and sprays is recommended only where ants become a domestic nuisance, threaten productivity or reach Other islands in the Seychelles group.

A scattergram method was developed for explaining the variability in the levels of fat, residual blood-meals and residual dry weights (RDW) of field samples of Glossina. The pattern of changes in mean RDW and fat levels was shown to be similar for five species of Glossina (G. morsitans morsitans Westw., G. swynnertoni Aust., G. fuscipes fuscipes Newst., G. palpoiis palpalis (R.-D.) and G. tachinoides Westw.) but different for the two sexes. On the basis of previous laboratory and field work, a simple descriptive model is developed that suggests priority pathways during the life of the adult fly for the products of successive blood-meals. Initially, for tenerals of both sexes, the priority is to increase thoracic musculature (RDW) without excessive production of fat. Subsequently, male flies devote each blood-meal exclusively to the production of fat, an energy store used mainly for flight activity. Females show a gradual increase in both fat and RDW levels, which probably corresponds with the development of the larva in utero. Wing-fray measurements support this interpretation. During each pregnancy cycle, females accumulate a reserve of fat over and above that necessary for the production of each larva. This is apparently required at the end of each cycle when res piratory losses exceed the energy supply from blood-meals taken. Mean thaematin levels of the four larger species studied (G. morsirans, G. swynnertoni, G. fuscipes and G. palpails) also showed a consistent pattern of change during each pregnancy cycle. Females apparently feed once at the beginning of each cycle, once at a stage of pregnancy when the larva in utero is approaching the moult from the second to the third instar and finally during the middle of the third larval instar, when the meal may be quite small and repeated.

Published laboratory measurements of the levels of fat and haematin in Glossina following a blood-meal were used to develop a prediction of the relationship between fat and haematin during the course of the feeding cycle. The fat-haematin relationship appears to be nearly linear over the later part of the feeding cycle. Data from three field studies onG. morsitans Westw.,G. swynnertoni Aust. and G. fuscipes Newst. were analysed in the way suggested by the prediction, and the observed rates of fat depletion were used to estimate flight activity of male flies towards the end of each feeding cycle. Results for G. palpalis (R.–D.) and G. tachinoides Westw., obtained in other studies with new electric sampling devices as well as by men with hand nets, illustrated particularly clearly the fat-haematin relationship over the whole of the feeding cycle. This was made possible by the electric traps catching more flies nearer to the start of their feeding cycle than the men with hand nets. A comparison between the different areas suggested that the fat level may trigger the onset of feeding behaviour inGlossina.

Components of dried citrus pulp were isolated and tested for arrestant activity for Atta cephalores (L.) and Acromyrmex octospinosus (Reich). Both lipid and non-lipid components contributed to the acceptability of the pulp. Arrestant activity was greater in the non-lipid fraction. Sucrose was found to be the most active single component of the non-lipids. The arrestant properties of the pulp were attributable to the particular balance of constituents rather than to any single component. The results are compared with the response of other insects to similar phytochemical stimuli.

Altering the proportions by weight of first-stage workers in groups of large workers from 26 to 40%, or of fourth-and fifth-stage reproductive nymphs in groups of mixed workers from 0 to 15%, did not affect wood consumption by laboratory cultures of Nasutitermes exitiosus (Hill). However, increasing the proportion of soldiers in groups of mixed workers from 0 to 15% decreased food consumption, which was very closely related to: the weight of workers present. The presence of 7·5% of large workers near to ecdysis had no significant effect, but food consumption was depressed when the proportion reached 15%, apparently due to high mortality among the moulting workers. It is concluded that caste composition can be an important factor in laboratory tests of materials against termite attack and that, unlike soldiers or large workers near to ecdysis, first-stage workers and late-stage reproductive nymphs of N. exitiosus are substantially independent castes, at least in their feeding.

A total of 2609 blood-fed females of 23 species of Aedes and Culex was collected, mainly from outdoor resting sites or in unbaited suction traps, at two localities in The Gambia. The host-feeding patterns of the mosquitoes indicated by the results of precipitin tests on the blood-meals were of thiee types: those in which large mammals were the predominant hosts, those where mammals, birds and reptiles were represented and those in which avian feeds predominated. C. tritaeniorhynchus Giles had fed predominantly on cattle, and small numbers of nine species of Aedes had fed on large mammals. C. thalassius Theo. was non-specific in its feeding habits, with feeds from a wide range of mammals and birds and a few reptiles. No evidence was found for a change in the feeding habits of C. thalassius with increasing age. C. poicilipes (Theo.) also fed on mammals and birds. Birds were the primary hosts of C. tigripes Grp. Charm., C. nebulosus Theo./ C. cinereus Theo., C. decens Theo., C. invidiosus Theo. and C. neavei Theo., although the feeds of C. decens included a large proportion on bats.

In an attempt to define the favoured hosts of Culex portesi Senevet & Abonnenc and C. taeniopus D. & K., one bat, ten rodents, four marsupials, two birds and one crustacean, one amphibian and one reptile were exposed in groups of six in suction traps located in a forest in Trinidad. Each group of six baits included two white mice as a standard. The numbers of mosquitoes caught at each bait on three or six replicate exposures were analysed by an analysis of variance to separate population and site variations. Baits were then compared for ‘attractiveness’ by ranking according to an adjusted mean catch for each mosquito species. The proportion of each mosquito species that engorged on individual baits was also determined. In some cases, attraction and engorgement varied widely, and this was attributed to defensive action on the part of the bait animal. It was concluded that C. taeniopus was particularly attracted by the opossum Didelphis, which was superior to all indigenous mammals, while C. portesi was attracted to a wide range of rodents and marsupials. Neither was attracted by the cold-blooded animals. Both mosquitoes engorged well on the relatively insensitive marsupials with long bare tails, Dideiphis, Marmosa and Caluromys, and also on the well-furred but diurnal squirrel. Of the mainly nocturnal remainder, Culex portesi showed higher engorgement rates than C. taeniopus, suggesting a more aggressive or persistent feeding behaviour.