Studies on Culicoides spp. biting midges in south-west Scotland in 1991 revealed that Culicoides impunctatus Goetghebuer, the dominant species throughout the season (May-September), was bivoltine, with a generation period of six weeks. Unseasonal climatic conditions during much of June resulted in a low proportion of first generation females surviving to the parous stage and a reduction in the overall, expected size of the second generation. Distinct behavioural phases of the female midges, such as resting, dispersal and host-seeking, were shown to be directly related to parity states (determined by fat body condition and ovarian development) during the gonotrophic cycle. Female flight activity increased after laying the first, autogenous egg batch, whereas males never flew far from the emergence site. The mean parous rate formula was used to calculate the survival rate of second generation nulliparous females, which was 0.57 per gonotrophic cycle.

Anagyrus mangicola Noyes, a primary parasitoid of the mealybug Rastrococcus invadens Williams, a pest of fruit crops in West Africa, was studied under controlled conditions at 27°C and 75% r.h. Data are given on the larval stages, their morphology and duration of immature development. The sex ratios of emerging parasitoids, adult longevity, host stage preference, the relationship of the size of the emerging parasitoid to host stage and development time are also investigated.

Pupae and larvae from various recorded hosts of the cecidomyiid Asphondylia gennadii (Marchal), a pest of carob on Cyprus, are found to be anatomically similar. These are redescribed and illustrated. The species appears to be a Palaearctic generalist species analogous to the Nearctic Asphondylia websteri Felt. Asphondylia capsici Barnes (Syn. n.) is considered a new junior synonym of A. gennadii. An unnamed soybean pod gall midge in Japan appears also to be identical to A. gennadii.

The predacious mite Euseius addoensis addoensis (McMurtry) is an important predator of citrus thrips, Scirtothrips aurantii Faure, and Tetranychid mites on citrus in the eastern Cape Province of South Africa. It can also survive and reproduce on certain pollens. In this study, pollen collected from windbreaks commonly used around citrus orchards, was compared with that from Carpobrotus muirii (Mesembryanthemaceae) as a food source for predacious mites. Bioassays were conducted in the laboratory using a leaf-disk technique and both mite mortality and fecundity were recorded through one generation. Casuarina cunninghamiana (Casuarinaceae) was the only windbreak tested that produced pollen capable of sustaining E. a. addoensis for a generation with fecundity similar to that obtained with Carpobrotus muirii. Pollens collected from Pinus radiata (Pinaceae), Cupressus arizonica, C. goveniana and C. torulosa (Cupressaceae) were unsuitable. Casuarina spp. windbreaks may therefore help to maintain predacious mite populations during autumn when natural sources of prey are diminishing.

Glossina morsitans morsitans Westwood and G. pallidipes Austen were caught hourly on electric nets from dawn to dusk in the Zambesi Valley, Zimbabwe, in cold and hot weather. When middle of the day temperatures reached only 24°C, activity of G. morsitans was at a low level from dawn to early afternoon but then rose to a dusk peak (most marked in males); G. pallidipes behaved similarly, but showed almost no activity before noon. When dawn temperatures rose above ca. 24°C (and afternoon temperatures above 38°C) the evening peak disappeared, to be replaced by a peak at dawn (and with almost no activity occurring through the rest of the day). This reversal of the more usual field activity pattern of late afternoon maxima reveals that the positive correlation between activity level and temperature at dusk (Brady & Crump, 1978) breaks down above ca. 33°C, although the earlier conclusion that the typical U-shaped activity pattern of tsetse flies is mainly driven by endogenous timing is still valid for moderate temperatures.

Choice between prey species by Chilocorus nigritus (Fabricius) and C. bipustulatus (Linnaeus) adults and larvae was determined. Strong preferences were absent and differences in the predators' feeding histories, were not reflected in their choices. The deleterious effects of a prey substitution, during larval development and adult maintenance, were investigated using C. nigritus with supplementary work on C. bipustulatus. These diet changes significantly retarded larval development rate, and subsequent adults were smaller than control individuals. Prey substitutions in the adult diet, suppressed oviposition for several days and feeding rate was reduced at one day after substitution. Larvae were more sensitive than adults to such diet changes. This was not a case of classical unsuitability of the new prey, as the fecundity and feeding rate returned to the same levels as before the substitution after a few days of exposure to the new prey. Furthermore, both prey types were suitable for larval development of C. nigritus when they fed on one exclusively, but unsuitable when substitutions were made. The effects of prey substitutions may possibly be attributed to the presence, in the new prey, of plant toxins, which the predators are initially not physiologically capable of dealing with in large quantities. These results present difficulties for the concepts of monophagy and polyphagy, being less well defined than normally thought. In view of coc-cinellid foraging behaviour and larval habitat selection by adults, the temporary reduction in fitness following a diet change is considered to be adaptive.

Adults of Autographa gamma (Linnaeus) make regular, seasonal migrations into areas where they are unable to breed continuously. Individuals migrate into Britain each spring, and after one, two or three generations, offspring of the spring migrants return to over-wintering areas in North Africa and the Middle East. The larvae are highly polyphagous and have been recorded damaging many crops, especially Brassica spp. Insects usually migrate during the adults' pre-reproductive period (PRP). The length of this period is therefore an index of migratory potential since individuals with longer PRPs have more time to express their capacity for flight and to travel further. Individuals reared outside in Britain in autumn have longer PRPs than individuals reared outside in summer. Laboratory experiments show that PRPs are extended at low temperatures and under short photoperiods and there is some evidence that PRPs are also extended under gradually decreasing photoperiods. By extending PRPs in response to environmental conditions typical of Britain in autumn, individuals are able to increase their potential for pre-reproductive migratory flight and therefore their potential for reaching suitable overwintering sites further south.

A seed-feeding curculionid, Rhyssomatus marginatus Fåhraeus, was introduced into South Africa for biological control of a perennial leguminous weed, Sesbania punicea, from South America. R. marginatus has become established even though its larval food source, seeds in ripening pods, is severely depleted by another biological control agent, an apionid, Trichapion lativentre (Beguin Billecocq), which destroys nearly all the flower-buds produced by the plants and reduces seed-set by 98% on average. In spite of this indirect competition, R. marginatus destroys up to 88% of the remaining seeds that develop on S. punicea, regardless of host plant location. Detrimental intraspecific interactions between larvae of R. marginatus are minimized, and the utilization of a depleted food source is optimized, through meticulous selection of oviposition sites by the female weevils. Although numerically inferior, R. marginatus is an important supplementary agent to T. lativentre and together the two species of weevils have almost arrested the reproductive potential of S. punicea.

Eight steers of European breed, Bos taurus type which were shown to be negative for antibodies against Theileria parva, were divided into two groups of four animals each. Animals in one group were inoculated with 0.5 ml undiluted tick-derived T. p. parva Muguga strain and animals in the other group were inoculated with 1 ml undiluted tick-derived T. p. parva Kilae strain to infect them. The two infected groups of cattle were simultaneously infested with uninfected nymphal Rhipicephalus appendiculatus Neumann in separate cloth patches. Ticks were from five populations, i.e., four from different geographical zones in Kenya and one laboratory population, in separate cloth patches. After moulting, the adult ticks were fed on rabbits for three days and their salivary glands were examined by microscopy for infective stages of the parasite. This revealed significant differences in the five populations as regards to their susceptibility to Theileria parva parasites.

The oviposition, larval arrest and establishment by the stem borer Chilo partellus (Swinhoe) were measured on different maize genotypes during anthesis. The maize genotypes tested were: Inbred A (Susceptible), Mp 704, MBR-8637, MBR-8650, MBR-8668, Poza Rica 7832, ER-29 SVR, Katumani Composite B, MMV 400, Bulk CG 4141, and ICZ2-CM. The moths oviposited on the middle leaves of the plant and the neonate larvae moved to leaf sheaths and ear husks to feed (arrest). As they grew older, the larvae invaded the stem, ear shanks and tassel to complete feeding (establishment) and to pupate. The response of the stem borer differed with resistant and susceptible maize genotypes at anthesis. More eggs were laid and larval arrest was higher on the susceptible than on resistant genotypes. Genotypes also differed in the resistance of feeding sites to larval establishment; the genotype Poza Rica 7832 showed resistance to larval establishment in the stems and ears, but the shank was heavily infested. By contrast, MBR 8637 showed greater resistance to larval establishment in the shank than the other genotypes. These resistance sources could be utilized effectively in a maize breeding programme to develop varieties with moderate to high levels of resistance at all potential feeding sites.

Trypanosome infections were monitored in three species of tsetse fly (Glossina pallidipes Austen, G. morsitans centralis Machado, and G. brevipalpis News-tead) at four locations in the Kagera River region of Rwanda from May 1989 to September 1990. Two of the four areas (Mpanga Ranch and Bukora Ranch) were subjected to tsetse fly suppression operations with odour-baited traps. Proboscis infections of the Trypanosoma congolense and T. vivax types accounted for roughly equal numbers of the 207 mature infections detected (3.8%). Variation in infection rates was area-specific rather than tsetse species-specific. Order of magnitude differences in tsetse fly densities among areas were not correlated with differences in infection rates at the start of tsetse fly suppression operations. Similarly, declines in population density on both control and experimental areas were not associated with significant changes in infection rates. The prevalence of trypanosomiasis in cattle at Bukora Ranch was not affected by a roughly 90% reduction in Glossina densities. T. congolense accounted for 79% of the infections at an overall prevalence rate of 5.5%. Trypanosomiasis in cattle persisted at extremely low densities of about 0.1 fly/trap/day. Treatment of cattle with diminazene aceturate (BereniR) suggested that many T. congolense parasites were drug resistant, and hence, were cycling among cattle due to the few Glossina present.

Shin Etsu twist-tie ropes containing the spiny bollworm (Earias insulana(Boisduval)) pheromone were tested for mating disruption in a cotton field in Israel during 1991. Two applications of the Shin Etsu rope formulation (1000 ropes/ha), containing 50 mg pheromone per rope, effectively disrupted mating of E. insulana, reduced damage to cotton bolls, and minimized the number of chemical treatments needed. The success of the mating disruption, which lasted for more than 3.5 months, was indicated by the following: 1. Reduction of trap catches in the pheromone-treated plot as compared with the control plot (commercially treated with insecticides). 2. Lower infestation of cotton bolls in the pheromone-treated as compared with the control plot. 3. Over 90% reduction in mating success in the pheromone-treated plot, as indicated by the number of mating pairs collected at night in the two plots. 4. Reduced number of insecticidal treatments applied in the pheromone-treated plot. The release rate of the pheromones was monitored by periodical gas chromatography analyses of the remaining pheromone in ropes. The half lives of the E. insulana and Pectinophora gossypiella (Saunders) (Lepidoptera: Gelechiidae) formulations were similar, with t1/2=45-60 days. In Israel, E. insulana and P. gossypiella infest cotton fields almost simultaneously and are controlled by the same insecticides. Therefore, the control of E. insulana with pheromones is essential in order to give reasonable meaning to the commercial application of pheromones for the control of P. gossypiella, as is practised today in many cotton fields. The present study indicates that control of E. insulana by the mating disruption technique is viable.

A key is provided for the recognition of previously described Encarsia species which are known to attack the tobacco whitefly, Bemisia tabaci (Genna-dius). 28 nominal species are included, of which type material has been examined for 22. Eight new synonymies are proposed: Encarsia aleyrodis (Mercet) and E. partenopea Masi are both synonymized with E. inaron (Walker); E. bemi-siae (Ishii) is synonymized with E. transvena (Timberlake); E. citri (Ishii) is synonymized with E. strenua (Silvestri), E. angelica Howard and E. deserti Rivnay & Gerling are both synonymized with E. luteola Howard; E. perspicuipennis (Girault) is synonymized with E. quaintancei (Howard); E. tabacivora Viggiani is synonymized with E. pergandiella Howard. E. hispida De Santis is removed from synonymy with E. meritoria Gahan. A lectotype is designated for E. porteri (Mercet). The known distributions and alternative hosts of Encarsia species attacking B. tabaci are provided as well as references to biology and notes on their use in biological control.

Larval responses of Delia radicum (Linnaeus), D. floralis (Fallén) and D. antiqua (Meigen) to plant volatiles were measured using a behavioural assay. Larvae were attracted to their host-plant, but non host-plants either had no effect or were repellent. Testing of individual secondary plant volatiles indicated that larvae were attracted to secondary plant volatiles from their host-plant, but were neither attracted nor repelled by secondary plant volatiles from non host-plants. A general plant volatile did not have this effect.

The potential distribution of the South African weevil, Perapion antiquum (Gyllenhal), a biological control agent for the weeds Emex spp., was deter mined by the computer program CLIMEX, using its native distribution, phenology and abundance together with development parameters. The predicted distribution included parts of Hawaii where the weevil successfully controlled Emex australis and E. spinosa. In Australia, sites of past unsuccessful releases have climates that this analysis indicates are unsuitable for the insect. The most favourable regions for establishment of the weevil are near the coast in the southern half of Australia, but most of these do not overlap with regions where Emex spp. are a problem. In western Cape Province, South Africa, E. australis plants are abundant and the weevil attacks the plant after seeds have formed. In Hawaii, a fortuitous combination of climatic conditions favours the weevil during the period after seed germination, and this may be the key to its control of the weed. Sites with climatic conditions similar to successful control sites in Hawaii are not found in Australia. It was concluded that P. antiquum will be of limited use as a biological control agent in Australia even in areas suitable for its establishment.

Phenotypic and genotypic variation was investigated between biotype E clones of the greenbug, Schizaphis graminum (Rondani). S. graminum was collected from wheat and sorghum fields in seven Kansas counties, USA. Colonies representing single maternal lineages (clones) were initiated and maintained in an insectary, and were verified as biotype E. An allozyme survey among clones did not detect any polymorphism at 24 enzymatic loci. We were unable to detect interclonal variation for nymphal survivorship, age at first reproduction, fecundity, and reproductive lifespan. Variation among genetically homogeneous daughters within clones was as great as variation among different clones for the above parameters, due to microenvironmental effects. However, by using the intergenic spacer in the rDNA cistron as a molecular fingerprinting tool, we identified 56 unique maternal lineages among 61 clones tested. The length of the intergenic spacer varied within individual S. graminum, and different maternal lineages had unique sets of spacer sizes. However, S. graminum within a single clone shared the same sets of spacer sizes. Extensive molecular genetic variation was found among biotype E clones, despite a lack of variation in additive genetic traits. Thus, biotype E populations are made up of many genetically distinct clones and do not represent an asexual proliferation of one or a few maternal lineages.

Biconical, F3, and wind oriented (WOT) traps, and black cloth targets, baited with the odour attractant swormlure-4, were assessed as catching and killing devices for the New World screwworm fly, Cochliomyia hominivorax (Coquerel), in Mexico. Among the traps, the WOT was significantly better than the others, with a mean catch 2.7 and 86.4 times better than those of the biconical and F3, respectively. Release of swormlure-4 could be reduced from the standard 10 ml'day to 2 ml'day without a reduction in the numbers caught in a WOT. Use of electric nets demonstrated that a large visual target was not necessary for the precise location of a swormlure-4 source by C. hominivorax. Target colour was important with respect to the landing response on targets: in a two-choice situation, flies landed much more frequently on black than on blue or yellow, and more on these two colours than they did on white. C. hominivorax tend not to circle a target before landing on it: about 75% of the flies caught on a combination of electrified black target plus electric flanking net were caught on the target. The potential for use of insecticide-impregnated cloth targets in suppression of adult C. hominivorax is discussed.

Iridomyrmex humilis (Mayr) and some other ants, through preference, efficient foraging and recruitment, quickly removed artificially exposed batches of insect eggs including those of the serious pest of eucalyptus, Phoracantha semipunctata (Fabricius). This supports evidence of the importance of some ants as predators of the many insects that lay freely exposed eggs. However, P. semipunctata normally lays eggs in peeling bark crevices where the approximately 60% laid within gaps less than ca. 0.65 mm wide could not be reached even by small ants such as I. humilis and Pheidole pallidula (Nyland-er). In tree trunks, which can become heavily infested after felling, egg predation by ants seems unimportant compared with the later large mortality of competing larvae, so, egg predation cannot account for the observed striking positive correlation between presence of I. humilis and lack of Phoracantha semipunctata damage to standing trees. Egg predation and other effects of I. humilis may be more important on living trees which are attacked initially by relatively few P. semipunctata. This needs more detailed investigation as do the overall egg predatory roles of ant species in biological control.