Fourteen populations of the western flower thrips Frankliniella occidentalis Pergande, originating from different hosts and countries in Asia, Europe, North America and New Zealand, were analysed for their competency and efficiency to transmit tomato spotted wilt virus (TSWV). All populations acquired and subsequently transmitted the virus, and were thus competent to transmit. They show marked differences in their efficiency, expressed as the percentage of transmitting adults. Efficiencies varied from 18% for a F. occidentalis population from the USA (US2) to 75% for a population from Israel (IS2). The differences between populations were not affected by the amount of virus ingested or by the host plant used. However, the tospovirus species studied and age at which the larvae acquired the virus affected the efficiency to transmit. First instar larvae of the NL3 population from The Netherlands were able to acquire tomato spotted wilt virus, whereas second instar larvae failed to do so. However, both instars of this population acquired impatiens necrotic spot virus (INSV), another tospovirus. This and tomato spotted wilt virus were both acquired by both larval stages of the populations IS2 and US2, although their ability to acquire virus decreased with their age. Hence, it is likely that, in general, both instar larvae of most F. occidentalis populations are competent to acquire both tospoviruses. These results show that large differences exist in the efficiency by which tomato spotted wilt is transmitted by the various F. occidentalis populations and that the ability to acquire tospovirus decreases with the age of the larvae

A parasitoid community on agromyzid leafminers from Cordoba, Argentina was analysed in terms of parasitoid assemblage size and host ranges of parasitoid species. Samples were taken during 1991–1995 at natural, urban and agricultural habitats. The system consisted of 69 parasitoid species and 51 leafminer species on 109 plant species. On average, 12 species parasitized each host, when only numerically well represented leafminer species were considered for analysis. Each parasitoid species exploited on average seven species and three genera of agromyzid leafminers. Both parasitoid assemblage size and parasitoid host range were notably high when compared with similar systems from other regions.

Trichogramma spp. egg parasitoids are generally absent in eggs of Helicoverpa armigera (Hübner) collected from chickpea, Cicer arietinum. In this study, the plant characters responsible for the absence of egg parasitoids and the feasibility of increasing parasitism levels on chickpea by mass-releasing Trichogramma chilonis Ishii were investigated. The residence time of female T. chilonison chickpea leaves was affected by trichomes and the acidic trichome exudates secreted on all green parts of the plant. The parasitoids spent a longer time on chickpea leaves where the acidic trichome exudates had been washed off than on unwashed leaves, and longer on leaves of a glabrous chickpea mutant than on washed leaves. When placed on unwashed chickpea leaves, 6.8% of the parasitoids were trapped and killed by the exudates. In a filter paper bioassay, female T. chilonis were deterred by high concentrations of malic and oxalic acids, the major components of the trichome exudate. Acetone and hexane extracts from the surface of chickpea leaves did not elicit a response from the parasitoids in the bioassay. Similarly, the parasitoids did not respond to volatiles emitted by chickpea plants in a four-armed airflow olfactometer. No parasitized eggs were collected from a chickpea field in which T. chiloniswere released five times at a weekly interval at a rate of > 137,000 females ha–1. Sticky trap catches showed that no parasitoid population was sustained in the release field.

Host range experiments and field sampling in Nigeria produced no evidence that African rice gall midge, Orseolia oryzivora Harris – Gagné, can develop on plants other than Oryza species. Sampling in three outbreak areas during 1994 showed that the insect's annual cycle varied according to the agroecological zone and rice cropping pattern. In the humid forest zone, Orseolia oryzivora persisted through the short dry season on ratoons of cultivated rice Oryza sativa at a rainfed site and on dry season rice crops at an irrigated one. In contrast, at rainfed sites in the moist savannah zone the pest survived the longer dry season on the perennial wild rice O. longistaminata, while ratoons and volunteers of O. sativa provided 'bridges' between the wild host and wet season rice crops. Early in the wet season at rainfed sites, galls of Orseolia oryzivora were not found at high density on wild rice, ratoons or volunteers. The heavy infestations which developed by October resulted primarily from rapid multiplication on rice crops themselves during the wet season. At all 13 sites sampled, the large majority of galls were found on fallow or cropped rice fields, rather than in ditches, bunds or uncultivated wetland, irrespective of the time of year or the hosts involved. From gall dissections, the parasitoids Aprostocetus procerae (Risbec) and Platygaster diplosisae Risbec caused over 30% mortality at some sites by October but generally increased too late to prevent crop damage. Implications of the results for the management of Orseolia oryzivora are discussed.

Development rates of the aphid, Brachycaudus rumexicolens (Patch), a recent arrival in Australia and a potential biological control agent against weeds in the family Polygonaceae, were measured over a range of constant temperatures. The theoretical lower limit for development is 6.4°C and the upper limit 32°C. Maximum fecundity per day was reached at 19°C. The rate of increase peaked at about 28°C giving a population doubling time of less than two days. These values were used with the current distribution to develop a CLIMEX model to predict the potential world distribution of the aphid. The model predicts that the aphid has suitable periods of population growth in autumn and spring, and that survival is unlikely over summer in most of south-western Australia where the aphid has the potential to contribute to the biological control of the polygonaceous weeds, Emex and Rumexspecies. The model predicts that years with cool summer temperatures and late summer rains in south west Australia, such as in 1990 when the aphid was first abundant, will be particularly suitable for aphid development. These conditions occurred twice between 1985 and 1995. To increase the effectiveness of the aphid as a biological control agent of weeds in other years, augmentation by provision of alternative hosts and/or the release of mass reared individuals during autumn is proposed.