Laboratory experiments were conducted to measure the susceptibility of larvae of the sandfly Phlebotomus papatasi Scopoli to the insect pathogen Bacillus sphaericus applied to the larval breeding medium. B. sphaericus had no effect on larvae aged 30 ± 2 days old, but larvae aged 13 ± 2 and 20 ± 2 days were highly susceptible and a clear dose response effect was detected. When B. sphaericus was left in the rearing medium for 18 days there was a significant increase in larval mortality, indicating persistence and possibly recycling. The results support the feasibility of employing B. sphaericus in the integrated control of sandflies.

Monitoring of Glossina austeni Newstead populations in the forested areas of Unguja island of Zanzibar has since November 1990 routinely been carried out with the sticky panel trap because commonly used tsetse traps (Biconical, Epsilon, F3) have proved to be unsuccessful in catching G. austeni. Initial studies on the catching ability of various types of sticky panels for G. austeni indicated that the monopanel was as efficient in catching flies as the 3-dimensional version and the smaller legpanel. No significant differences in catch rate and sex ratio were observed with monopanels in various colours and colour combinations. Legpanels coloured white on one panel side and blue on the other side caught significantly more flies compared with other colour combinations, but female flies were under-sampled (32%). The type of sticky material applied on the panel influenced significantly the catch rate and female ratio. During long-term trapping with baby blue and white monopanels, females were under-sampled (38–46%) except when polyisobutyleneLMW was used as sticky material. Analysis of the age composition of the sampled G. austeni females revealed that tenerals and nulliparous were well represented (11–24%). More than 20% of the trapped females were older flies, i.e. females with four or more ovulations, but this percentage dropped to 10% when Tanglefoot was used as sticky material.

The development of palatable baits for the suppresion of pest termites relies on combining phagostimulants with slow-acting toxicants that termite foragers carry back to the colony and distribute to nest-mates. In the present study the palatability and toxic effects of three compounds, ampicillin, tetracycline and urea to the subterranean termites Reticulitermes flavipes (Kollar) and R. virginicus (Banks) were investigated. 1.0% solutions (w/v) of ampicillin and tetracycline applied to filter paper were unpalatable to termites in choice tests, and these concentrations depressed termite feeding, survivorship, individual biomass, and numbers of gut protozoa in no-choice tests. In contrast, urea solutions of 0.1%, 1.0% and 5.0% (w/v) were palatable to termites in choice tests, but 5.0% urea solutions depressed termite survivorship and protozoan numbers in no-choice tests. Addition of urea to tetra- cycline solutions resulted in increased palatability and decreased termite survivorship over tetracycline administered alone. The toxic effects in combination with its palatability to Reticulitermes spp. indicate that urea may be a promising candidate for termite control programmes.

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 response of populations of the saw-toothed grain beetle, Oryzaephilus surinamensis (Linnaeus) (7022/1 strain), to insecticide treatments of different extent and standard concentration was studied over 16 weeks (no more than two generations) in experimental bins containing a wheat bulk and 16 kibbled wheat-filled refuges. Populations comprised initially low frequencies of resistant phenotypes, as would occur in natural populations where resistant alleles had newly arisen. The degree of control achieved was related to the extent of treatment. Distribution of adult beetles was greatly influenced by the effects of insecticide treatment superimposed on intrinsic microhabitat preferences. Live beetles were recovered from all bins, although numbers sampled from wheat bulks of treated bins were 1–6% of that of the untreated bin. More live beetles were sampled from wheat bulks than were recovered from all respective refuges, except in one treated bin. Live beetles were more numerous in refuges in corners and in those furthest from refuges on treated walls. Beetles were more likely to die in refuges in coiners and lower down in the bins or in refuges on treated walls adjacent to untreated walls. Although treatments produced heavy mortality, the frequencies of resistant individuals did not increase significantly. This could either be because treatments were equally successful in eliminating homozygous susceptible and heterozygous individuals, or because selection in favour of resistant individuals was too weak to be detectable, especially if there were negative correlations between viability and fecundity. Significant genetic structuring of refuge subpopulations with respect to the resistance phenotypes was noted and appeared to be related to behavioural preferences for microhabitat.

The ecology of Microtermes subhyalinus Silvestri in savanna and an adjoining sugarcane plantation in the Central African Republic is described. Other species of termites were associated with savanna and sugarcane but M. subhyalinus was the most prevalent and the only one causing damage to sugarcane. Damage to the crop by M. subhyalinus was assessed throughout the year. Termite attacks were confined initially to the root system just after cutting but as the cane matured and growth ceased, termites tunnelled up into the above-ground stems. Losses accounted for 5–10% of total annual production. The nesting and foraging behaviour of this species altered with the change from natural savanna to sugarcane cultivation. The vertical distribution of M. subhyalinus fungus combs in soil under savanna and sugarcane over different seasons was analysed. There was a linear relationship between monthly rainfall and foraging activity in savanna while in sugarcane no relationship was observed.

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.

Predatory feeding on Tetranychus urticae Koch (Acari: Tetranychidae) populations on cotton by phytophagous thrips, Thrips imaginis Bagnall, T. tabaci Lindeman and Frankliniella schultzei Trybom (Thysanoptera: Thripidae), was investigated in the field and laboratory. Phytophagous thrips are a common early season pest of cotton in Australia, though their true pest status is undefined. In California, the phytophagous thrips (Frankliniella occidentalis (Pergande)) is regarded as an opportunistic predator of mite eggs, their consumption of which increases fitness over a diet of leaf tissue alone. Thrips are among the most abundant of insects on young cotton. If they consume mite eggs, even at relatively low rates, they could have a significant influence on the probability of survival of early season spider mites. Consumption of eggs of T. urticae by thrips was investigated in the laboratory. Second instar F. schultzei consumed more eggs per day (ca. 4 eggs per day) than did second instar T. tabaci or T. imaginis (ca. l egg per day). Consumption by first instar F. schultzei was much lower than for second instars. Adult T. tabaci consumed ca. l egg per day whilst adults of F. schultzei consumed only ca. 0.5 eggs per day, although some individuals of this species did consume substantial numbers of eggs. Larvae of all thrips species showed a type II functional response to prey density. In the field, adults and larvae of T. tabaci and F. schultzei showed a preference for cotton seedlings that were also infested by spider mites. In a glasshouse, larvae of T. tabaci showed a highly significant preference for feeding within mite colonies. In the field, suppression of predators, predominantly T. tabaci and F. schultzei, with a broad spectrum insecticide (dimethoate) contributed to outbreaks of mites occurring earlier than they would have otherwise. The results show that phytophagous thrips eat mite eggs and that they are potentially important predators of spider mites in the field, especially given their abundance on young cotton and preference for inhabiting situations in which mite colonies are found.

Yellow starthistle (Centaurea solstitialis Linnaeus) (Compositae), native to Eurasia and the Mediterranean Basin, is a spiny and poisonous, naturalized, invasive weed of grasslands and other environments in the western USA. Chaetorellia australis Hering is a trivoltine natural enemy of yellow starthistle whose larvae feed on developing seeds within capitula. Following host-specificity studies in Greece and the USA, C. australis was first introduced from Greece into the USA in 1988 for the biological control of yellow starthistle. Field samples of yellow starthistle capitula revealed that 10.3% (four years after initial release) and 18.0% (three years after initial release) of capitula at Colfax, Washington, and Merlin, Oregon, respectively, were infested by ≥ 1 C. australis (mature larvae or empty puparia). Both sites also contained naturalized, weedy populations of, Centaurea cyanus, a second host-plant of Chaetorellia australis. The fly was also recovered from Centaurea cyanus capitula two years after its release at Kendrick, Idaho. The potential importance of the presence of C. cyanus for Chaetorellia australis and biological control of yellow starthistle is discussed. At the Colfax site, all infested yellow starthistle capitula contained 1 C. australis, while at the Merlin site 28.8% of infested capitula contained > 1 C. australis, with a mean of 1.36 C. australis per infested capitulum. A preliminary seed destruction impact study at the Merlin site showed that there was a 83.4% reduction in seeds per capitulum infested by ≥ 1 C. australis, and the presence of > 1 C. australis resulted in a 95.4% reduction in seeds per infested capitulum.

The Palaearctic pine aphids, Eulachnus agilis (Kaltenbach) and Eulachnus rileyi (Williams) have both been introduced into other continents where they have been reported causing damage to economically important pines. In Euorpe, they are attacked by the specialist parasitoid Diaeretus leucopterus (Haliday) which has been suggested as a possible biological control agent. Here we report on several aspects of the ecology of the parasitoid, conducted on E. agilis in Germany in 1993–94, to provide a more scientific basis for judging its potential for use in biological control. Parasitism of all instars in the field rarely exceeded 10% and was independent of host density. A high percentage of parasitized aphids were hyperparasitized. Measurements of the impact over 16 weeks in a greenhouse-release experiment showed that parasitism rates increased from 2 to 19% but were insufficient to suppress the aphid population below a damaging level. There was evidence of a density-dependent response. Studies on foraging behaviour showed that female parasitoids searched pines by quite extensive walks. There was no relationship between the number of aphids per tree and number of ovipositions, and the mean number of ovipositions per female per tree was 2.5 ± 0.4 eggs. The majority (55.1%) of encountered aphids did not respond to the parasitoids. However, female parasitoids attacked a much higher proportion of aphids that did respond but oviposition success on this group was poor. The number of aphid-infested needles on seedlings increased significantly owing to the parasitoid's foraging activity. On the basis of these results, it is suggested that D. leucopterus is only likely to be of benefit in biological control if used in conjunction with other complementary natural enemies.

Factors (developmental state, starvation time, host local density and host distance from insects) affecting the response of the ragwort flea beetle, Longitarsus jacobaeae (Waterhouse), to upwind plants of tansy ragwort, Senecio jacobaeae (Compositae), were examined in a wind tunnel. There was seasonal variation in beetle response to host odours. Individuals collected during spring and early summer (6 May to 14 July 1993) showed directed movement toward the upwind plants. However, individuals collected in mid-summer (23 July 1993) showed no significant response to upwind host plants and walked randomly in the wind tunnel. Individuals collected during late summer to autumn (4 August to 23 October 1994) once again showed response to upwind plants, whereas those collected in winter (January 1994) walked randomly in the wind tunnel. The absence of beetle response to plant odours coincided with potential summer aestivation and winter ‘hibernation’. For spring and early summer beetles that responded to plant odours, the strength of the response did not change significantly with the number of plants (1, 2, 4 or 6) presented upwind in the tunnel nor with the starvation time (2, 6, 10, 12, 24 or 36 h) of the tested beetles, and these patterns were consistent for male and female beetles. Individuals responded to plants from a distance of 60 and 300 cm. The speed of movement, similar for males and females, increased slightly as L. jacobaeae oriented more directly toward host plants.

We compare the pest status and age structure of bush fly,Musca vetustissima Walker, populations, and the abundance of native dung beetles at Uriarra, ACT, for five fly seasons before (1976–81) and three seasons after (1990–93) the exotic dung beetles Euoniticellus fulvus (Goeze) and Onthophagus taurus (Schreber) became established in the area during the late 1980s. Fly populations in all post-introduction seasons of 1990–93 were much older than in the pre-introduction seasons, indicating reduced levels of local fly breeding and higher proportions of immigrants. During the pre-introduction period, the fly annoyance index exceeded the ‘minimum discomfort’ threshold of 10.0 in parts of all seasons. The mean index was 11.3, and this did not differ significantly between seasons. The total dry weight of dung beetles at the time of the first influx of bush flies each year was consistently below 3 g/dung pad which was apparently too low to depress bush fly breeding below their replacement level of 3%. The mean annoyance index of 3.1 for the 1990–91 season was significantly lower than for the pre-introduction period and the total dry weight of beetles during November-December exceeded 3 g/dung pad on all sampling occasions. In the 1991–92 and 1992–93 seasons fly annoyance indices reverted to pre-introduction levels (seasonal means of 13.7 and 7.8 respectively) and total dry weight of beetles was again below 3 g/dung pad when the bush flies first appeared. Survival of immature bush flies in local dung pads remained below the replacement level of 3% during 1990–91, but exceeded 3% on two occasions during 1991–92. Apart from 1976–77, when mean catches of native dung beetles (O. australis and O. granulatus) were exceptionally high, mean catches of native dung beetles during pre- and post-introduction years were very similar, but the mean dung beetle biomass increased following the establishment of exotic species.

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.

Spider mites, Tetranychidae, represent one of the most cosmopolitan and economically important groups of terrestrial arthropods; however, many aspects of their evolutionary relationships remain uncertain. We sequenced part of the mitochondrial cytochrome oxidase subunit I (COI) gene in 20 species of phytophagous mites belonging to nine genera and two families (Tetranychidae and Tenuipalpidae), including several agricultural pests. As eported in insects, the sequences were extremely rich in A + T (75% on average), especially in the third codon position (95%). However, one of the genera we studied had a significantly lower A + T content (69% on average, 78% in the third codon position), showing that base composition can change substantially over short periods of time. Most interspecific differences were transversions and their number increased steadily with the number of non-synonymous differences, while the number of transitions remained constant. The phylogeny based on COI sequences was inferred using the maximum likelihood method. The results are compatible as a whole with the traditional classification based on morphological characters, but call for some minor taxonomic revisions. Some morphological characters and life history traits (mode of reproduction, adaptation to the host plant) were also analysed within this phylogenetic framework. At the family level, one can see a trend towards thelytoky becoming rarer compared to the general mode of reproduction of the group, arrhenotoky. There is also an evolutionary tendency towards a more complex mode of life, with the production of silk webs and correlated changes of the locomotion apparatus. However, in the Tetranychidae there seems to have been convergent evolution of these morphological characters together with independent development of a common adaptation to this mode of life in different genera.