Aerial dispersal using silk (‘ballooning’) has evolved in spiders (Araneae), spider mites (Acari) and in the larvae of moths (Lepidoptera). Since the 17th century, over 500 observations of ballooning behaviours have been published, yet there is an absence of any evolutionary synthesis of these data. In this paper the literature is reviewed, extensively documenting the known world fauna that balloon and the principal behaviours involved. This knowledge is then incorporated into the current evolutionary phylogenies to examine how ballooning might have arisen. Whilst it is possible that ballooning co-evolved with silk and emerged as early as the Devonian (410–355 mya), it is arguably more likely that ballooning evolved in parallel with deciduous trees, herbaceous annuals and grasses in the Cretaceous (135–65 mya). During this period, temporal (e.g. bud burst, chlorophyll thresholds) and spatial (e.g. herbivory, trampling) heterogeneities in habitat structuring predominated and intensified into the Cenozoic (65 mya to the present). It is hypothesized that from the ancestral launch mechanism known as ‘suspended ballooning’, widely used by individuals in plant canopies, ‘tip-toe’ and ‘rearing’ take-off behaviours were strongly selected for as habitats changed. It is contended that ballooning behaviour in all three orders can be described as a mixed Evolutionary Stable Strategy. This comprises individual bet-hedging due to habitat unpredictability, giving an underlying randomness to individual ballooning, with adjustments to the individual ballooning probability being conferred by more predictable habitat changes or colonization strategies. Finally, current methods used to study ballooning, including modelling and genetic research, are illustrated and an indication of future prospects given.

Cosmopolites sordidus (Germar) is an important pest on bananas and plantains. Population build-up is slow and damage becomes increasingly important in successive crop cycles (ratoons). Yield loss results from plant loss, mat disappearance and reduced bunch size. Damage assessment requires destructive sampling and is most often done on corms of recently harvested plants. A wide range of damage assessment methods exist and there are no agreed protocols. It is critical to know what types of damage best reflect C. sordidus pest status through their relationships with yield loss. Multiple damage assessment parameters (i.e. for the corm periphery, cortex and central cylinder) were employed in two yield loss trials and a cultivar-screening trial in Uganda. Damage to the central cylinder had a greater effect on plant size and yield loss than damage to the cortex or corm periphery. In some cases, a combined assessment of damage to the central cylinder and cortex showed a better relationship with yield loss than an assessment of the central cylinder alone. Correlation, logistic and linear regression analyses showed weak to modest correlations between damage to the corm periphery and damage to the central cylinder. Thus, damage to the corm periphery is not a strong predictor of the more important damage to the central cylinder. Therefore, C. sordidus damage assessment should target the central cylinder and cortex.

The pea aphid Acyrthosiphon pisum Harris has been shown to produce an increasing proportion of winged morphs among its offspring when exposed to natural enemies, in particular hoverfly larvae, lacewing larvae, adult and larval ladybirds and aphidiid parasitoids. While these results suggest that wing induction in the presence of predators and parasitoids is a general response of the pea aphid, the cues and mechanisms underlying this response are still unclear. Tactile stimuli and the perception of chemical signals as well as visual signals are candidates for suitable cues in the presence of natural enemies. In this paper the hypothesis that the aphids' antennae are crucial for the wing induction in the presence of natural enemies is tested. Antennae of pea aphids were ablated and morph production was scored when aphids were reared either in the presence or the absence of predatory lacewing larvae over a six-day period. Ablation of antennae resulted in a drastic drop in the proportion of winged morphs among the offspring, both in the presence and the absence of a predator whereas predator presence increased wing induction in aphids with intact antennae, as reported in previous experiments. The results show that antennae are necessary for wing induction in the presence of natural enemies. Critical re-examination of early work on the importance of aphid antennae and tactile stimuli for wing induction suggests that a combination of tactile and chemical cues is likely to be involved not only in predator-induced wing formation but also for wing induction in response to factors such as crowding in the aphid colony.

Infection of Lacanobia oleracea(Linnaeus) larvae with the microsporidium Vairimorpha necatrix(Kramer) resulted in significant effects on the survival and development of the braconid parasitoid, Meteorus gyrator(Thunberg). Female M. gyrator did not show any avoidance of V. necatrix-infected hosts when they were selecting hosts for oviposition. When parasitism occurred at the same time as infection by the pathogen, or up to four days later, no significant detrimental effects on the parasitoid were observed. However, when parasitism occurred six to eight days after infection, a greater proportion(12.5–14%) of hosts died before parasitoid larvae egressed. Successful eclosion of adult wasps was also reduced. When parasitism and infection were concurrent, parasitoid larval development was significantly faster in infected hosts, and cocoons were significantly heavier. However, as the time interval between infection and parasitism increased, parasitoid larval development was significantly extended by up to two days, and the cocoons formed were significantly (c. 20%) smaller. Vairimorpha necatrix spores were ingested by the developing parasitoid larvae, accumulated in the occluded midgut, and were excreted in the meconium upon pupation.

Choice of the targeted host developmental stage, regulation of parasitoid numbers released and introduction of food supplements are operational factors with a potential to influence the level of biological control. In a closed laboratory storage system maintained over two generations of the host, the impact of these three parameters on the control potential of the parasitoid Dinarmus basalis Rondani was investigated for high populations of larvae of Acanthoscelides obtectus (Say) feeding inside dry common bean seeds Phaseolus vulgaris. The beans were already infested with immature bruchids at the beginning of the storage period to simulate harvest conditions, characterized in a previous study. Treatments resulted in a reduction of 48–75% of the bruchid population within 16 weeks of storage. The best timing of parasitoid release was at the simulated harvest, as later releases reduced the bruchid population only by about half this percentage. Host feeding is postulated to be the key factor involved in the observed difference. The effect of increasing the number of parasitoids strongly depended on host age and food supplement. Addition of vials with honey had no direct effect on the bruchid population or on the parasitoid progeny. The ecological significance of these findings and implications for biological control are discussed.

Young plants of Sitka spruce, Scots and Corsican pine were subject to high and low light, and high and low nitrogen treatments in a polyhouse experiment. The effect of treatments on resin duct size and nitrogen concentration in stem bark was determined together with feeding by Hylobius abietis Linnaeus on the stems of ‘intact’ plants and on ‘detached’ stems cut from the plant. Resin duct size was largest on Corsican pine and smallest on Sitka spruce and inherent variation in duct size between the three conifer species appears to determine the pattern of weevil feeding between species. Resin ducts and the flow of resin from them protect the stems of young conifers from weevil feeding not by affecting the total amount of bark eaten but by limiting the depth of feeding and so protecting the inner phloem and cambium. Shallow feeding may increase the likelihood of effective wound repair. Duct size was positively related to plant growth and in particular increased with bark thickness. Overall, ducts were largest in the high light treatment although species differed in their response to the treatment. It is suggested that the effects of plant size, growing conditions and transplantation on susceptiblilty to attack by H. abietis, reported in various studies, may be due to underlying variation in resin duct size or flow rate. The effect on weevils of super-ficial feeding on stems is to increase the time for reproductive maturation by reducing consumption of the inner bark which has a higher nitrogen content.

The drive towards a more sustainable and integrated approach to pest management has engendered a renewed interest in conservation biological control, the role of natural enemy communities and their interactions with prey. Monoclonal antibodies have provided significant advances in enhancing our knowledge of trophic interactions and can be employed to help quantify predation on target species. The tetragnathid spider Pachygnatha degeeri Sundevall was collected from fields of winter wheat in the UK and assayed by ELISA for aphid proteins. It was demonstrated that this spider did not simply consume greater quantities of aphids because it was bigger. In addition, P. degeeri contained significantly greater concentrations of aphid in their guts than other spiders, showing that aphids comprised a greater proportion of their diet. Although P. degeeri constituted only 6% of the spider population numerically, females and males respectively contained 16% and 37% of total aphid proteins within all spiders screened, significantly more than their density would predict. These spiders also preyed upon aphids at a disproportionately high rate in June, during the aphid establishment phase, theoretically the best time for limiting growth in the aphid population. Although less abundant than other generalist predators, the capability of these hunting spiders to consume large numbers of aphids highlights them as a more significant component of the predator complex than had previously been realized. Limitation of aphid numbers early in the year by generalist predators provides more time for the specialist aphid predators and parasitoids to move in.

This study was conducted in the humid forest zone of Cameroon, in 2002 and 2003. The main objective was to investigate the effects of intercropping on infestation levels and parasitism of the noctuid maize stem borer Busseola fusca Fuller. Two trials were planted per year, one during the long and one during the short rainy season. Maize monocrops were compared with maize/legume or maize/cassava intercrops in two spatial arrangements: maize on alternate hills or in alternate rows. Spatial analyses showed that the stemborer egg batches were regularly dispersed in the maize monocrop and aggregated in the intercrops, as indicated by b, the index of dispersion of Taylor's power law. Depending on the crop association and planting pattern, intercrops reduced the percentage of plants with stem borer eggs by 47.4–58.4% and egg densities by 41.2–54.5% compared to monocropped maize. Consequently, larval densities were 44.4–61.5% lower in intercrops compared to monocrops. Intercropping maize with non-host plants did not affect larval parasitism. Up to two-fold higher levels of egg parasitism by scelionid Telenomus spp. were recorded in inter- compared to monocrops during the short rainy seasons of 2002 and 2003. No differences were found among the mixed cropping treatments and parasitism was lower during the long compared to the short rainy seasons. It was proposed that differences in levels of parasitism were due to density dependence effects rather than the effect of the presence of non-host plants in the system.