The biology, damage and control of the sugarcane pest Pyrilla perpusilla Walker are reviewed. The present systematic position, distribution and the range of alternative host plants are surveyed and the life cycle and the extent of damage caused by the pest are considered in detail. The factors influencing the abundance of the insect are discussed, with emphasis on their use in reducing the population of the pest. Much of the published data in this area is based on un replicated observations. However, potential biological control agents for the insect are evaluated with special reference to the most effective nymphal and adult ‘parasitoid’, the moth Epiricania melanoleuca (Fletcher). Many other potential biological control agents have received some study, but in most cases, little more than basic biological information on phenology and life cycle has been published. As considerable changes in chemical control practices over the past fifty years have occurred, these, together with other control methods such as agronomic, mechanical, cultural and host plant resistance approaches are discussed along with prospects for the future, sustainable control of the pest. Research areas which are in need of more work are identified: these are biological control and cultural practices. It is suggested that an integrated approach to future research should be made, incorporating at least host-plant resistance and predation.

Helicoverpa armigera (Hübner) larvae were collected from field crops and wild hosts in India, Nepal and Pakistan from 1991 to 1995, and ninety eight laboratory cultures established. Cypermethrin, fenvalerate, endosulfan, quinalphos, monocrotophos and methomyl insecticides were topically applied to 30–40 mg, first laboratory generation larvae and resistance determined from log dose probit bioassays. Significant levels of cypermethrin and fenvalerate resistance were found in all field strains, demonstrating that resistance to at least some pyrethroids is now ubiquitous in H. armigera populations in the Indian subcontinent; cypermethrin and fenvalerate resistance levels ranged from 5– to 6500–fold and 16– to 3200–fold respectively. Pyrethroid resistance levels were highest in the intensive cotton and pulse growing regions of central and southern India where excessive application of insecticide is common. In all field strains assayed, pre-treatment with the metabolic synergist piperonyl butoxide (pbo), resulted in significant suppression of pyrethroid resistance. However, in nearly all cases, full suppression of resistance was not achieved. This residual non-pbo-suppressible resistance was most likely due to a nerve-insensitivity resistance mechanism. Pbo-insensitive resistance was highest in regions of India where insecticides were frequently applied to cotton and legume crops. In some regions where insecticides were heavily overused, a second high order nerve-insensitivity mechanism (possibly a Super -Kdr type mechanism), may have been present. Incipient endosulfan resistance (1–28-fold), was present throughout India, Nepal and Pakistan. Low to moderate levels of resistance (2–59–fold), were reported to the phosphorothionate group organophosphate, quinalphos, in India and Pakistan, but there was no evidence of significant resistance (0.4–3–fold), to the phosphate group organophosphate, monocrotophos, under our bioassay conditions between 1993 and 1994. H. armigera strains collected in Nepal in 1993 and 1994 were susceptible to quinalphos, but by 1995, 4–5–fold resistance was detected. It is probable that much of the resistance to pyrethroid, organophosphate and carbamate insecticides in the Indian subcontinent can be attributed to an inherited or inducible mixed function oxidase complex. Non-pbo-suppressible resistance becomes significant in regions and periods in the season when insecticide selection pressure on resistant H. armigera larvae on cotton and legume crops is very high.

Single applications of slow-release pheromone formulations were used to control yellow stem borer, Scirpophaga incertulas Walker, by mating disruption in two trials on rice in India. The first trial compared the efficacy of two formulations containing 1:10:1 and 1:10:0 ratios of (Z)-9-hexadecenal, (Z)-ll-hexadecenal and (Z)-9-octadecenal, components of the S. incertulas pheromone, and a commercially-available formulation, Selibate CS, containing the related Chilo suppressalis Walker pheromone, a 1:10:1 blend of (Z)-9-hexadecenal, (Z)-ll-hexadecenal and (Z)-13-octadecenal, with farmers' practice plots treated with insecticides. Pheromone trap catches indicated that in each of the pheromone-treated plots the catches of male moths were reduced by up to 98% compared with catches in the insecticide-treated plots suggesting that pheromone-mediated communication was disrupted. Larval damage ranged from 5.7 to 8.1% white heads (WH) in the insecticide-treated plots compared to a significantly reduced range of 2.1 to 2.4% WH in the pheromone-treated plots. There was no significant difference between the damage estimates recorded for each of the different pheromone treatments. A second trial compared the effect on efficacy of applying the two-component pheromone formulation on different dates, 9–12 and 39–44 days after transplanting. Both pheromone treatments gave significantly lower levels of WH damage compared to equivalent insecticide treated plots; 1.8, 2.0, and 6% respectively. In both trials rice grain yields increased by between 5 and 12% in pheromone-treated plots compared to insecticide treated plots, which was due, at least in part, to the cumulative effect of reduced stem borer damage in both the tillering and reproductive stages of the crop cycle. The formulations exhibited pseudo-zero order release rates with field lives of between 70 and 90 days.

Helicoverpa armigera (Hübner), H. assulta Guenée and Heliothis peltigera (Denis & Schiffermüller) were collected as mixed populations from safflower and the wild host Datura metel, from Patancheru, Andhra Pradesh, India, in 1992 and 1993, and their toxicological responses to insecticides determined. Both Helicoverpa armigera strains were highly resistant to cypermethrin, fenvalerate, endosulfan and quinalphos insecticides, based on resistance ratios relative to laboratory reared susceptible strains. There was no evidence of resistance development in H. assulta and Heliothis peltigera to the same chemicals. Light trap data collected from 1974 to 1987 showed that Helicoverpa armigera was at least 100 × more abundant than the other two species over most of the cropping season. Peak catches of H. assulta and Heliothis peltigera were confined to defined times in the season, corresponding with the flowering and fruiting periods of their respective host plants; August-October for Helicoverpa assulta and November-December for Heliothis peltigera. Helicoverpa armigera on the other hand, because of its high polyphagy on commercial and wild hosts, was abundant between August and April. Resistance has not developed in H. assulta and Heliothis peltigera in southern India, probably because of their restricted host range, limiting exposure to insecticides.

Salivary oxidation has been hypothesized as a possible means by which spotted alfalfa aphid (Therioaphis trifolii maculata (Buckton)) deals with deleterious phenolics in the susceptible host plant, lucerne (Medicago sativa). In this study, some basic questions about this hypothesis were examined. Constitutive flavonoids and a crude phenolic extract from lucerne leaves were shown to be deterrent to the aphid in a choice test. Phenolics were oxidized by phenoloxidase in the saliva of the aphid. Fresh sap expressed from aphid-infested, oxidatively-reactive leaves was phagostimulatory to the aphid, compared to that from non-infested healthy leaves. Oxidative reaction was responsible for at least part of the phagostimulation, since addition of ascorbic acid into the sap deleted the phagostimulatory activity. Ascorbic acid was itself not deterrrent and its titre decreased in leaves after aphid feeding. The results provide further information on the Oxidative interaction between spotted alfalfa aphid and lucerne plants and support the previous suggestion that oxidation of phenolics and anti-oxidation by reductants affects feeding activity of this aphid.

Field strains of housefly collected from animal units were exposed to the insecticide baits, Golden Malrin and Alfacron in laboratory cage tests. Flies were exposed to each bait alone in no-choice tests and to the bait plus sugar in choice tests. Golden Malrin was generally effective against the strains tested in no-choice tests; the lowest recorded knockdown after exposure for 48 h was 82%. Eight strains, however, gave a reduced knockdown when exposed in choice tests (between 60 and 87%). Golden Malrin was ineffective against one strain in choice tests with 7% knockdown at 48 h. With Alfacron 12 strains exhibited knockdown responses between 57 and 87% at 48 h in no-choice tests. In choice tests, 17 strains gave knockdown responses of less than 50% at 48 h. Correlation coefficients were calculated for results from cage tests and levels of resistance to the active ingredients assessed using topical application and feeding test methods. There was a positive correlation between cage and feeding test results. Correlation was, however, less significant with choice test results than with no-choice results. There was no correlation between cage test and topical results. It is suggested that behavioural resistance was responsible for the reduced responses shown by some strains in choice tests compared to no-choice tests. The results demonstrate the importance of using a laboratory method which allows for behavioural differences when monitoring for resistance to insecticides formulated as baits. Behavioural differences in housefly strains should be considered when testing the efficacy of new bait formulations. It is also suggested that insecticide baits should not be used continuously and exclusively to control houseflies in intensive animal units.

Using regression analysis the early season dynamics of Helicoverpa punctigera (Wallengren) were determined from long series of light trap catches (10–19 years) from three sites in Australia (Narrabri and Trangie in New South Wales, and Turretfield in South Australia). The size of the second spring generation (G2, the one causing major pest problems in summer cropping regions) was strongly related to the size of the first spring generation (G1). In most cases, rainfall in early winter had a positive influence on the size of G2, whereas rainfall in spring had a negative effect. Regressions were found to account for 49 to 93% of the annual variation in G2, depending on site. The use of light trap catches and weather data to forecast pest levels from a few months to a few weeks in advance is discussed, along with the improved understanding of early season H. punctigera dynamics.

Chlorpyrifos was applied to field lucerne at a rate (0.3 kg/ha) recommended to control the weevil Sitona discoideus Gyllenhal. The fate of the weevil population and that of the immature stages of its parasitoid Microctonus aethiopoides Loan were examined by field sampling. A single application of chlorpyrifos reduced the weevil population by 96.2% 14 days after application. There was no significant decline in the control population between sampling intervals nor a change in the parasitism levels. However, in the insecticide-treated plots c. 88% of survivors (n=8) were parasitized. The effects of chlorpyrifos residues at 0.3 and 1.0 kg/ha were also examined. At the lower rate, residues produced c. 40–60% weevil mortality for up to 15 days after application although a significant decline in activity only occurred between days 1 and 3. At 1.0 kg/ha, the activity decay curve was much slower with no significant decline in weevil mortality for the first 15 days of the experiment. Weevil gender was shown to be a significant factor influencing mortality. Mortality of unparasitized weevils was also significantly greater than for parasitized hosts while the stage of parasitoid development had a significant effect on mortality. Where parasitoid development was sufficiently advanced, late-instar larva emerged and successfully pupated from cadavers up to 18 h following host mortality. The consequences of applying insecticide to control S. discoideus as they relate to the parasitoid and the success of biological control are discussed.

Descriptions are given of the larval stages of the 15 most damaging African stemborer species, belonging to the genera Chilo Zincken, Coniesta Hampson, Eldana Walker, Maliarpha Ragonot, Scirpophaga Treitschke (Pyralidae), Sesamia Guenée and Busseola Thurau. Chaetotaxy of the larvae has been studied. Previously published studies on the same subject are discussed. A key is provided for the identification of the genera, and where possible, the species.

Two small-scale experiments were carried out to test the effectiveness of narrow target barriers against re-invasion by Glossina pallidipes Austen and G. m. morsitans Westwood. The barriers consisted of either one or two lines of targets, with the targets placed at distances of c. 16 m apart within the lines. At this spacing the targets are so close together that a fly could scarcely fail to see at least one target as it crossed the barrier. Nonetheless, in both cases, flies penetrated the barriers with probability c. 10% as judged by catches in odour-baited traps on either side of the barrier. Narrow barriers, even at very high target density, are ineffective against tsetse invasion. This accords with previous theoretical predictions and mark-recapture studies, and suggests that wider barriers are needed for the prevention of re-invasion of cleared areas by tsetse.

Marked male moths of Plutella xylostella (Linnaeus) released at up to 12 m from a sticky delta trap baited with synthetic pheromone, were caught in significantly higher numbers than moths released at greater distances, both in downwind and upwind directions, from the traps. Maximum catches were obtained when the traps were set at 30 cm above the crop level. Continued exposure of septa in the field led to a progressive reduction in catches, and catches were significantly reduced after 28 days. Newly emerged virgin female moths attracted significantly more male moths than one day old or older virgin female moths. Five day old virgin females attracted no male moths. When virgin females were used as a bait in the field, the peak catch of males was between 1600 and 1800 h. Tests at two field locations revealed that, at the concentration tested, synthetic sex pheromone attracted a similar number of moths to the number attracted by 10 virgin females in delta traps, and significantly more than 8, or fewer virgin females.

The success of the chrysomelid beetles, Chrysolina hyperici (Forster) and C. quadrigemina (Suffrian), used widely for biological control of St John's wort (Hypericum perforatum) seems to depend on the synchronization of their phenologies with climate. The factors governing the termination of diapause in the field and laboratory were assessed in New Zealand. Before summer diapause, C. hyperici made up more than 70% of adult beetles in the field. Both species entered summer diapause at the beginning of January, but C. quadrigemina terminated diapause 3–4 weeks earlier (early March) than C. hyperici. Both male and female C. quadrigemina and male C. hyperici adults became sexually mature during summer diapause but C. hyperici females still had immature ovaries. Females of C. quadrigemina began laying eggs immediately after summer diapause but females of C. hyperici did not oviposit until more than 2 weeks after they terminated diapause in late March. In the laboratory summer diapause was terminated in both species by short day length, not by simulated rainfall, contradicting previous assumptions that autumn rainfall is the critical factor. The demonstrated differences in phenology are reflected in different overwintering strategies for the two species. Because C. hyperici overwinters in the egg stage, this species will be a more successful biological control agent of St John's wort in countries with cold winters than C. quadrigemina, which overwinters in the more vulnerable larval stage. However, in countries with mild winters C. quadrigemina is more successful in controlling St John's wort than C. hyperici, since it has a developmental lead with its larvae feeding and growing during the winter.

The effects of maize–cowpea intercropping on three lepidopteran stemborers (Chilo partellus (Swinhoe) C. orichalcociliellus (Strand) and Sesamia calamistis Hampson) and their natural enemies were studied in Kenya. Oviposition was not affected by intercropping, but significantly fewer larvae and pupae were found in the intercrop. Egg parasitism was significantly higher in the intercrop reaching as high as 80% and was positively density dependent. Larval and pupal parasitism remained between 3 and 20%, but neither of these were affected by intercropping or host density. The main predators, ants and wandering spiders, were not augmented by intercropping, but an inverse relationship in abundance was found between these two predator groups. It is concluded that maize intercropped with cowpea has only limited potential as a method of controlling the key pests in maize.

In Zimbabwe, studies were made of the responses of Glossina pallidipes Austen to known and candidate repellents. Various chemicals, dispensed at c. 5–10 mg/h, were placed beside Epsilon traps already baited with a blend of acetone, octenol, 4-methylphenol and 3-n-propylphenol. Pentanoic or hexanoic acids or acetophenone halved the catch and 2-methoxyphenol reduced the catch by 90%. There were no consistent differences in the responses of males and females. Pentanoic acid or acetophenone or 2-methoxyphenol at an unbaited trap reduced the catch by 40%, 75% and 60%, respectively. Baiting traps with a combination of pentanoic acid, acetophenone and 2-methoxyphenol did not reduce the catch below that produced by 2-methoxyphenol alone. Pentanoic acid and 2-methoxyphenol reduced the efficiency of traps from c. 40% to 20% but acetophenone had no significant effect. Acetophenone and 2-methoxyphenol halved the numbers of tsetse attracted to a target baited with acetone, octenol and phenols but none had a significant effect on the proportion that landed. 2-Methoxyphenol significantly reduced the numbers of tsetse attracted to a source of natural ox odour but only pentanoic acid had a significant effect on feeding responses, reducing the proportion that fed on an ox from 59 to 45%. It is concluded that these repellents do not provide any useful degree of protection against trypanosomiasis. In areas where tsetse are abundant (500 bites/ox/day) and infection rates in tsetse are high (>5%) it is highly unlikely that any repellents would be useful. However, in areas where tsetse are less abundant (1 bite/ox/day) and/or infection rates are low (c. 0.5%) the potent, unidentified repellents present in human odour might effectively complement the control of disease using trypanocidal drugs.

The mealybug Maconellicoccus hirsutus (Green) occurs in many tropical and subtropical parts of the Old World and extends into some temperate areas. It has now reached the West Indies where it is causing extensive damage to plants. There is concern that it may be introduced to the southern USA and to Central and South America. A brief account is given of reported damage caused by M. hirsutus to some commercial and food crops in Asia and Africa and the species is redescribed. M. multipori (Takahashi), now known from a wide area in southern Asia, is redescribed, and M. ramchensissp. n. from Nepal, is described as new. These two species are closely related to M. hirsutus and can easily be mistaken for it. M. hirsutus and M. multipori are sometimes intercepted at quarantine inspection of plants and plant produce in the USA and Europe. In order to aid identification, a key is provided to the eight species presently included in the genus Maconellicoccus Ezzat.