Locating a host plant is crucial for a phytophagous (herbivorous) insect to fulfill its nutritional requirements and to find suitable oviposition sites. Insects can locate their hosts even though the host plants are often hidden among an array of other plants. Plant volatiles play an important role in this host-location process. The recognition of a host plant by these olfactory signals could occur by using either species-specific compounds or specific ratios of ubiquitous compounds. Currently, most studies favor the second scenario, with strong evidence that plant discrimination is due to central processing of olfactory signals by the insect, rather than their initial detection. Furthermore, paired or clustered olfactory receptor neurons might enable fine-scale spatio-temporal resolution of the complex signals encountered when ubiquitous compounds are used.

The roots of most land plants are colonised by mycorrhizal fungi that provide mineral nutrients in exchange for carbon. Here, we show that mycorrhizal mycelia can also act as a conduit for signalling between plants, acting as an early warning system for herbivore attack. Insect herbivory causes systemic changes in the production of plant volatiles, particularly methyl salicylate, making bean plants, Vicia faba, repellent to aphids but attractive to aphid enemies such as parasitoids. We demonstrate that these effects can also occur in aphid-free plants but only when they are connected to aphid-infested plants via a common mycorrhizal mycelial network. This underground messaging system allows neighbouring plants to invoke herbivore defences before attack. Our findings demonstrate that common mycorrhizal mycelial networks can determine the outcome of multitrophic interactions by communicating information on herbivore attack between plants, thereby influencing the behaviour of both herbivores and their natural enemies.

cis-Jasmone, or (Z)-jasmone, is well known as a component of plant volatiles, and its release can be induced by damage, for example during insect herbivory. Using the olfactory system of the lettuce aphid to investigate volatiles from plants avoided by this insect, (Z)-jasmone was found to be electrophysiologically active and also to be repellent in laboratory choice tests. In field studies, repellency from traps was demonstrated for the damson-hop aphid, and with cereal aphids numbers were reduced in plots of winter wheat treated with (Z)-jasmone. In contrast, attractant activity was found in laboratory and wind tunnel tests for insects acting antagonistically to aphids, namely the seven-spot ladybird and an aphid parasitoid. When applied in the vapor phase to intact bean plants, (Z)-jasmone induced the production of volatile compounds, including the monoterpene (E)-␤-ocimene, which affect plant defense, for example by stimulating the activity of parasitic insects. These plants were more attractive to the aphid parasitoid in the wind tunnel when tested 48 h after exposure to (Z)-jasmone had ceased. This possible signaling role of (Z)-jasmone is qualitatively different from that of the biosynthetically related methyl jasmonate and gives a long-lasting effect after removal of the stimulus. Differential display was used to compare mRNA populations in bean leaves exposed to the vapor of (Z)-jasmone and methyl jasmonate. One differentially displayed fragment was cloned and shown by Northern blotting to be up-regulated in leaf tissue by (Z)-jasmone. This sequence was identified by homology as being derived from a gene encoding an ␣-tubulin isoform.

Stem borers, comprising the larvae of a group of lepidopterous insects, and parasitic witchweeds, particularly Striga hermonthica and S asiatica, cause major yield losses in subsistence cereal production throughout sub-Saharan Africa. Studies are described that have led to the development of a`push-pull' strategy for minimising stem borer damage to maize and sorghum. This involved the selection of plant species that could be employed as trap crops to attract colonisation away from the cereal plants, or as intercrops to repel the pests. The two most successful trap crop plants were Napier grass, Pennisetum purpureum, and Sudan grass, Sorghum sudanensis. The intercrop giving maximum repellent effect was molasses grass, Melinis minuti¯ora, but two legume species, silverleaf, Desmodium uncinatum, and greenleaf, D intortum, gave good results and had the added advantage of suppressing development of S hermonthica. In terms of stem borer control, the plant chemistry responsible involves release of attractant semiochemicals from the trap plants and repellent semiochemicals from the intercrops. With M minuti¯ora, parasitism of stem borers was also increased by certain chemicals repellent to ovipositing adults. The mechanism of striga control has not been fully elucidated, but allelopathic effects from the Desmodium species have been shown to involve stimulation of germination and interference with haustorial development. Signi®cant bene®cial effects have been obtained with the individual components of these push-pull strategies. However, the most robust crop-protection package is obtained when these components are combined. The trap crop and intercrop plants also provide valuable forage for cattle, often reared in association with subsistence cereal production. There has been considerable take-up of the system within the communities where farmer-managed trials have been carried out, particularly in the Trans Nzoia and Suba districts of Kenya, and the programme is set to expand throughout and beyond Kenya.

Natural enemies respond to herbivore-induced plant volatiles (HIPVs), but an often overlooked aspect is that there may be genotypic variation in these 'indirect' plant defence traits within plant species. We found that egg deposition by stemborer moths (Chilo partellus) on maize landrace varieties caused emission of HIPVs that attract parasitic wasps. Notably, however, the oviposition-induced release of parasitoid attractants was completely absent in commercial hybrid maize varieties. In the landraces, not only were egg parasitoids (Trichogramma bournieri) attracted but also larval parasitoids (Cotesia sesamiae). This implies a sophisticated defence strategy whereby parasitoids are recruited in anticipation of egg hatching. The effect was systemic and caused by an elicitor, which could be extracted from egg materials associated with attachment to leaves. Our findings suggest that indirect plant defence traits may have become lost during crop breeding and could be valuable in new resistance breeding for sustainable agriculture.