Evidence of induced plant defences in a pondweed

SUMMARY. 1. Palatability changes, induced by herbivore damage, have received much attention in terrestrial plant-insect research but negligible investigation in freshwater systems.
2. Potamogeton coloratus (Potamogetonaceae) and the folivorous caddis larva Triaenodes bicolor (Insecta: Trichoptera) were used to duplicate bioassay techniques common in terrestrial plant-insect research.
3. Leaves of P. coloratus were artificially damaged. Grazing on damaged leaves, and undamaged leaves from damaged plants, was compared to that on control leaves from undamaged plants up to 10 days after initial damage.
4. The palatability of damaged and undamaged leaves from damaged plants declined compared to control leaves. The differences were significant at 5 days after damage, but had recovered to approximately equal palatability after 10 days.     

The ecological significance of rapid wound-induced changes in plants: insect grazing and plant competition

Summary This paper investigates the hypothesis that a rapidly induced phytochemical response to grazing damage, such as that seen in tomato, serves to deflect insect herbivores away from leaves soon after damaging them (the grazing dispersal hypothesis). As a result, grazing damage is more dispersed than it otherwise would be, and young leaves, which may be of particular importance to a plant in competition for light, are not damaged excessively. In the first experiment, artificial removal of c. 15% of leaf area led to a significant reduction in plant performance compared with undamaged controls, but only when the plants were grown together in competition for light. The second experiment demonstrated that the distribution of grazing damage within the plant was an important factor in the outcome of competition; in those plants in which grazing was applied to the lower leaves there was no effect of damage upon performance compared with undamaged controls, whereas grazing to the upper leaves significantly reduced plant performance. A third experiment provided some insight into how this interaction between damage and competition comes about. It was shown that damage to leaves led to a rapid drop in the rate of extension growth of the main shoot, especially when the upper leaves were damaged, and normal rates of growth were not resumed for at least 3 days. It is argued that in a rapidly growing canopy, such an effect may mean that a damaged plant loses its position in the height hierarchy. The final experiment showed that previous damage to plants can affect the distribution of subsequent grazing by larvae of Spodoptera littoralis, apparently through a wound-induced reduction in leaf palatability. Plants which had been artificially damaged 48 h previously were grazed significantly less than controls, and the avoidance effect was greatest in the young leaves. These results are consistent with the grazing dispersal hypothesis, and suggest that rapid wound-induced responses may be of greatest significance in species characteristic of fertile environments where competition for light is particularly intense.     

Variation in damage to cotton affecting larval feeding preference of Spodoptera littoralis

Feeding experiments with larvae of Spodoptera littoralis were performed with leaves from cotton plants subjected to damage and from undamaged plants. In the experiments, four different time intervals (1, 3, 7, and 14 days) after damage induction and two different levels (high and low) of herbivore damage were tested. Seven days after damage induction larvae fed less on the young top leaves from damaged plants for both levels of damage. At the high damage level, the larvae fed less on leaves from the damaged plants after just three days, and this effect still remained 14 days after damage infliction. When mature leaves from the middle of the plant were compared, no difference between treatments was observed.Two plant sizes were tested, small plants with 4–5 true leaves and large plants with 8–10 true leaves. In small plants the induced changes affecting larval feeding were found mainly in the youngest leaf at the top of the plant, while in large plants the induced effects were found in both the youngest and the second youngest leaves.In plants subjected to artificial damage, larvae fed less on top leaves of the damaged plants when compared to leaves from undamaged plants. When leaves from plants that had been artificially damaged were directly compared with leaves from plants damaged by herbivores, larvae fed more on the youngest leaves from artificially damaged plants when the plants were large. In small plants no significant difference was found when comparing artificial and herbivore damage.     

Pattern of defoliation and its effect on photosynthesis and growth of Goldenrod

1. Leaf area was removed from Solidago altissima in either a dispersed pattern (half of every leaf removed) or a concentrated pattern (every other leaf removed) and effects on leaf gas exchange, vegetative growth and flowering were examined relative to undefoliated controls. Gas exchange was measured for leaves remaining after defoliation and for regrowth leaves that developed post-damage (at 7, 16 and 26 days post-defoliation).
2. Area-based photosynthetic rates of leaves remaining after defoliation were not affected by either dispersed or concentrated damage, but damage of both types enhanced area-based photosynthesis of regrowth leaves at 16 days post-defoliation and to a lesser extent at 26 days post-defoliation.
3. Dispersed damage, but not concentrated damage, stimulated mass-based photosynthesis of undamaged leaves remaining after defoliation. Undamaged leaves remaining after defoliation and regrowth leaves on damaged plants had higher specific leaf area (leaf area/leaf mass) than comparable leaves on control plants. Mass-based photosynthesis was more strongly elevated by defoliation than area-based photosynthesis because of this increase in specific leaf area.
4. Plants with dispersed damage recovered more quickly from defoliation; they had higher relative growth rates in the first week post-defoliation than plants with concentrated damage. Both types of defoliation caused similar reductions in flower production.
5. These results add to accumulating evidence that dispersed damage is generally less detrimental to plants than concentrated damage and suggest that physiological changes in leaves may be part of the reason.     

Defoliation effects on isoprene emission from Populus deltoides

Isoprene emission from plants is one of the principal ways in which plant processes alter atmospheric chemistry. Despite the importance of this process, few long-term controls over basal emission rates have been identified. Stress-induced changes in carbon allocation within the entire plant, such as those produced by defoliation, have not been examined as potential mechanisms that may control isoprene production and emission. Eastern cottonwood (Populus deltoides) saplings were partially defoliated and physiological and growth responses were measured from undamaged and damaged leaves 7 days following damage. Defoliation reduced isoprene emission from undamaged and damaged leaves on partially defoliated plants. Photosynthetic rates and leaf carbon and nitrogen pools were unaffected by damage. Photosynthetic rate and isoprene emission were highly correlated in undamaged leaves on undamaged plants and damaged leaves on partially defoliated plants. There was no correlation between photosynthetic rate and isoprene emission in undamaged leaves on partially defoliated plants. Isoprene emission was also highly correlated with the number of source leaves on the apical shoot in damage treatments. Increased carbon export from source leaves in response to defoliation may have depleted the amount of carbon available for isoprene synthesis, decreasing isoprene emission. These results suggest that while isoprene emission is controlled at the leaf level in undamaged plants, emission from leaves on damaged plants is controlled by whole-branch allocation patterns. Received: 12 May 1998 / Accepted: 9 November 1998     

Opposing effects of spring defoliation on late season oak caterpillars

ABSTRACT. 1. The pedunculate oak, Quercus rohur L., suffers high annual levels of spring defoliation in Wytham Woods. near Oxford.
2. This spring defoliation affects late season caterpillars through a variety of damage-induced changes in the leaves.
3. Diurnea fagella (D. & S.), one of the commonest late season caterpillars, shows reduced larval survival and pupal weight on regrowth foliage when compared to undamaged primary foliage.
4. D. fagella also suffer higher larval mortality on naturally damaged primary foliage than they do on undamaged foliage.
5. Despite this, the three commonest late season caterpillar species are more abundant on damaged trees than undamaged ones. and their distributions are biased towards damaged leaves within the canopy.
6. Other factors that may be more important than leaf damage in determining the distribution and abundance of late season caterpillars are discussed. D.fagella larvae spin their larval refuges more rapidly on damaged than undamaged foliage, and this may reduce mortality by natural enemies, or ameliorate adverse effects of weather.     

Interactive effects of soil fertility and herbivory on Brassica nigra

Soil nutrient availability may affect both the amount of damage that plants receive from herbivores and the ability of plants to recover from herbivory, but these two factors are rarely considered together. In the experiment reported here, I examined how soil fertility influenced both the degree of defoliation and compensation for herbivory for Brassica nigra plants damaged by Pieris rapae caterpillars. Realistic levels of defoliation were obtained by placing caterpillars on potted host plants early in the life cycle and allowing them to feed until just before pupation on the designated plant. Percent defoliation was more than twice as great at low soil fertility compared to high (48.2% and 21.0%, respectively), even though plants grown at high soil fertility lost a greater absolute amount of leaf area (38.2 cm² and 22.1 cm², respectively). At both low and high soil fertility, total seed number and mean mass per seed of damaged plants were equivalent to those of undamaged plants. Thus soil fertility did not influence plant compensation in terms of maternal fitness. However, the pathways used to achieve compensation in seed production were different at low and high soil fertility. At low soil fertility, relative leaf growth rates (area added per inital area per day) of damaged plants were drastically reduced over the second week of caterpillar feeding. Damaged plants recovered the leaf area lost to herbivory in the two weeks following insect removal by increasing leaf relative growth rates above the levels seen for undamaged plants, but the replacement of leaf tissue lost to herbivory came at the expense of stem biomass. At high soil fertility, relative leaf growth rates of damaged plants were similar to those of undamaged plants both over the second week of caterpillar feeding and following caterpillar removal, and stem biomass was not affected by herbivory. These results suggest that higher levels of soil nutrients increased the ability of plants to stay ahead of their herbivores as they were being eaten. Because damaged plants at high soil fertility were able to maintain leaf growth rates to a greater extent than damaged plants at low soil fertility, they did not fall as far behind undamaged plants over the period of insect feeding and did not have as much catching up to do after feeding ended to compensate for herbivory.     

Prior flea beetle herbivory affects oviposition preference and larval performance of a potato beetle on their shared host plant

Abstract 1. The herbaceous plant Solanum carolinense (L.) (Solanaceae) is host to a number of specialist insects, including the leaf-feeding beetles Epitrix fuscula (Crotch) and Leptinotarsa juncta (Germar) (Coleoptera: Chrysomelidae). Potted individuals of S. carolinense were subjected to one of two treatments: exposure to herbivory by E. fuscula or exclusion of all herbivores. The effects of E. fuscula herbivory on larval performance and oviposition preference of L. juncta were investigated.
2. Although the masses of the L. juncta pupae did not differ between the two treatments, larvae feeding on damaged plants developed more slowly than those feeding on undamaged plants.
3. In both paired leaf choice trials and whole plant choice trials, larvae of L. juncta showed no preference for undamaged versus damaged hosts.
4. In a field transplant experiment, adult L. juncta females showed slight feeding preferences and strong oviposition preferences for undamaged plants versus plants that had been fed on by E. fuscula .
5. The results are discussed with reference to their implications for plant-mediated competition among herbivores and constraints on the evolution of plant resistance.     

Grazing tolerance of Gentianella amarella and other monocarpic herbs: why is tolerance highest at low damage levels?

Plants have adapted to compensate for the loss of vegetative biomass and reproductive potential caused by grazing. Shoot damage breaks down the correlative inhibition maintained by apical dominance. The consequent increased branching may lead to increased production of flowers and fruits in damaged plants, provided that enough resources, both in terms of meristems and nutrients, are available. In Gentianella amarella, the removal of the apex of the main stem (10% clipping) had no pronounced effect on branching and plant performance. In one of the two study populations, however, apically damaged plants produced more fruits than undamaged control plants. The plants also fully compensated for 50% removal of the main stem in terms of above-ground biomass, but their fruit production was reduced compared to control and apically damaged plants. After 75% clipping, fruit production was not significantly reduced compared to 50% clipping. Consequently, G. amarella showed highest tolerance in the presence of minor shoot damage. The pattern is qualitatively similar in some other monocarpic species (Gentianella campestris, Erysimum strictum and Rhinanthus minor). Multiple constraints as well as selective forces may shape these compensatory responses: (1) A lack of basal meristems may constrain tolerance of high damage levels. (2) Species with basal meristems may have a potential to tolerate major damage, but a shortage of resources or otherwise unfavourable growth conditions may constrain their compensatory ability. (3) It may be adaptive to have maximum tolerance of low and moderate damage levels if chemical defences reduce the risk of extensive shoot damage as well as the risk of repeated grazing. (4) The compensatory ability of monocarpic species may be affected by selective forces that favour fast vertical growth early in the season and unbranched architecture in undamaged conditions. Therefore, it is not the mere grazing history, but also other factors associated with growth conditions that are required to explain the variation in grazing tolerance.     

Effects of host plant leaf damage on cabbage looper moth attraction and oviposition

Mated femaleTrichoplusia ni (Hubner) moths, when presented a choice of either undamaged cotton plants,Gossypium hirsutum L., or damaged plants (cut leaves or feedingT. ni larvae) in a flight tunnel, were most often attracted first to the damaged plants. However, these same moths oviposited primarily on the undamaged plants. In a similar test with cabbage plants,Brassica oleracea L., the presence of conspecific larvae decreased both attraction and oviposition. Cuts to cabbage leaves had no significant effect on attraction or oviposition. When presented one plant at a time, percentages of cabbage looper moths attracted were not affected by the presence of larvae on either cabbage or cotton plants, or by cuts to cabbage plant leaves. Percentages of moths attracted were, however, higher using cotton plants with cut leaves. The results suggest an important role for damage induced plant volatiles in host location as well as host acceptance byT. ni.     

Towards an understanding of the mechanisms of tolerance: compensating for herbivore damage by enhancing a mutualism

Abstract.   1. Traditionally, losses in plant fitness or yield resulting from insect damage have been redressed by reducing pest populations using insecticides or biocontrol; these approaches rely on the untested assumption that reduced plant fitness or yield is caused by diminished resources available to damaged plants.
2. By experimentally manipulating pollination and damage levels independently, it is shown that pollination, as well as lack of resources, may be limiting to damaged plants in a model insect-pollinated crop, cantaloupe.
3. With enhanced pollination, damaged plants produce as much fruit as undamaged plants, even under high damage levels. In contrast, damaged plants without supplemental pollination produced significantly less fruit than undamaged plants.
4. This approach is unique in shifting the focus away from reducing pest populations and toward enhancing mutualistic interactions. It avoids risks posed by insecticides (which also kill pollinators) and by biocontrol agents, known threats to native species.
5. Determining the mechanism underlying compensation sheds light on recovery from insect damage in both natural and managed systems. These results have a bearing on managing native plant populations suffering from pollinator declines.
6. Finally, it may be predicted that resources could limit tolerance to herbivore damage in resource-poor or high competition environments, whereas pollination may limit tolerance when resource levels are high.     

Olfactory preferences of Popillia japonica, Vanessa cardui, and Aphis glycines for Glycine max grown under elevated CO2

Levels of atmospheric CO(2) have been increasing steadily over the last century and are projected to increase even more dramatically in the future. Soybeans (Glycine max L.) grown under elevated levels of CO(2) have larger herbivore populations than soybeans grown under ambient levels of CO(2). Increased abundance could reflect the fact that these herbivores are drawn in by increased amounts of volatiles or changes in the composition of volatiles released by plants grown under elevated CO(2) conditions. To determine impacts of elevated CO(2) on olfactory preferences, Japanese beetles (Popillia japonica Newman) and soybean aphids (Aphis glycines Matsumura) were placed in Y-tube olfactometers with a choice between ambient levels of CO(2) gas versus elevated levels of CO(2) gas or damaged and undamaged leaves and plants grown under ambient levels of CO(2) versus damaged and undamaged plants grown under elevated levels of CO(2). All plants had been grown from seeds under ambient or elevated levels of CO(2). Painted lady butterflies (Vanessa cardui L.) were placed in an oviposition chamber with a choice between plants grown under ambient and elevated levels of CO(2). A. glycines and V. cardui showed no significant preference for plants in either treatment. P. japonica showed no significant preference between ambient levels and elevated levels of CO(2) gas. There was a significant P. japonica preference for damaged plants grown under ambient CO(2) versus undamaged plants but no preference for damaged plants grown under elevated CO(2) versus undamaged plants. P. japonica also preferred damaged plants grown under elevated levels of CO(2) versus damaged plants grown under ambient levels of CO(2). This lack of preference for damaged plants grown under elevated CO(2) versus undamaged plants could be the result of the identical elevated levels of a green leaf volatile (2-hexenal) present in all foliage grown under elevated CO(2) regardless of damage status. Green leaf volatiles are typically released from damaged leaves and are used as kairomones by many herbivorous insects for host plant location. An increase in production of volatiles in soybeans grown under elevated CO(2) conditions may lead to larger herbivore outbreaks in the future.     

Insect herbivory in relation to dynamic changes in host plant quality*

Evidence that chemical changes in plants following insect feeding can lead to reduced grazing levels, enhanced insect movement and selective leaf avoidance is briefly reviewed. A simple model is constructed in which changes in damaged and/or adjacent leaves lead to effects on herbivore performance. The model reveals that as the density of herbivore larvae/plant increases from one to twenty-four, wound-induced changes in the leaves reduce larval survival by up to 40%, treble the number of movements of the larvae and increase their development time by c. 10%. The distribution of grazing between leaves changes in the direction of more leaves with lower grazing levels but overall grazing levels are not greatly affected by the above changes in larval performance. The model's output is discussed in relation to recent views concerning the relative roles of intra-specific competition and predation in regulating insect herbivore numbers.     

Beetle grazing reduces natural infection of Rumex obtusifolius by fungal pathogens

In previously reported laboratory experiments, infection of Rumex obtusifolius by the rust fungus Uromyces rumicis was decreased on leaves which had prior herbivory by the beetle Gastrophysa viridula . In this paper we investigate whether this interaction is robust for natural infection by a variety of fungi in field experiments carried out in spring and autumn with plants given different levels of nitrogen fertilization. Grazing by G. viridula led to a decrease in lesion density of Ramularia rubella and Venturia rumicis in the spring and V. rumicis and U. rumicis in the autumn experiment. For V. rumicis and U. rumicis significant reductions in lesion density occurred on the undamaged leaves of damaged plants, compared with similar leaves on undamaged plants, suggesting systemic induced resistance. This induced resistance was usually independent of the amount of nitrogen fertilization, although the inhibitory effect of grazing on R. rubella in the spring and V. rumicis in the autumn experiment was enhanced by increasing nitrogen fertilization and was inhibited by increasing nitrogen fertilization for V. rumicis in the spring. In both experiments, the lesion density of V. rumicis was greater on leaves on which R. rubella was also present, and the presence of U. rumicis in the autumn experiment was linked to a similar but greater effect on V. rumicis lesion density. We found no evidence of induced resistance by fungi against fungi in these experiments. We highlight the complex interactions between inhibitory and facilitatory processes acting on leaf fungal infection. These results are compared with the proposed molecular mechanisms of induced resistance(s) and we consider the benefits of closer integration between molecular and ecological investigations of induced resistances that occur in the field.     

Leaf damage by herbivores affects attractiveness to pollinators in wild radish, Raphanus raphanistrum

 We carried out two experiments to determine the effect of leaf damage on plant attractiveness to pollinators using wild radish, Raphanus raphanistrum (Brassicaceae), a self-incompatible annual herb. Pairs of plants from 36 full-sib families were grown in pots in the greenhouse. One member of each pair was damaged by Pieris rapae larvae that were allowed to remove half of the leaf area of each of the first four rosette leaves. The plants were subsequently taken out for pollinator observations once a week from the beginning of flowering in late June until the end of August. We conducted two experiments to examine how foliar damage affected visitation by pollinators. In the first experiment, numbers of pollinators visiting plants were compared between damaged and control sibling plants. In the second experiment, the number of open flowers during observations was controlled to be the same for both damaged and undamaged sibs. Damage significantly decreased the number and size of flowers during the first observations in late June. Damaged plants received fewer visits by native bees during the first week of observations. Since damage did not affect native bee visits when the number of open flowers was equalized between treatments, flower number was probably the main cue attracting native bees to plants. In the experiment without flower number control, syrphid flies, the other abundant pollinator taxon, spent more time per flower on the undamaged than on the damaged plants. When flower number was controlled, flies probed significantly more flowers during each visit on the undamaged than on the damaged plants and had higher visitation rates to undamaged plants early in the season. Since syrphid flies preferred undamaged plants both with and without flower number control, they apparently used cues apart from flower number for visitation. The difference between undamaged and damaged plants in floral characteristics and pollinator visitation vanished within a few weeks after the start of flowering. This result suggests that early damage may not have a strong fitness effect through reduction in mating success. However, poor weather conditions can cause early mortality of plants in the field, and nutrient depletion and competition decrease fruit set of later flowers. Therefore, conditions exist under which visitation to early flowers may affect plant fitness. Received: 30 July 1996 / Accepted: 10 February 1997     

Compensatory branching and changes in nitrogen content in the aquatic weed Salvinia molesta in response to disbudding

Summary Secondary side branching in Salvinia molesta plants grown in nutrient solution was dependent on the availability of nitrogen. The compensatory response of S. molesta to damage by complete manual disbudding was the production of extensive secondary side branching at nitrogen levels below those required for this type of branching in undamaged plants. After 28 days damaged plants had the same number of ramets as undamaged plants but the dry weight was considerably reduced.After 7 days there was no change in dry weight with partial or complete disbudding. However completely disbudded plants had no new ramets and plants with all terminal buds removed had numbers of ramets significantly reduced. The concentration of nitrogen in the tissue of new growth was higher in damaged than undamaged plants. The results are discussed in relation to the biological control of S. molesta by the bud-eating weevil, Cyrtobagous salviniae.     

Induced resistance of Gossypium australe against its most abundant folivore,Bucculatrix gossypii

Induced resistance of cultivated Gossypium to its exotic, agricultural pests is well studied but little is known about whether native cottons respond to damage by endemic herbivore populations. This study examined induced responses of Gossypium australe to its most abundant folivore, Bucculatrix gossypii. Prior damage did not affect the number of new mines initiated. Survival of miners on damaged, young leaves and cotyledons was reduced compared with survival on young leaves and cotyledons of undamaged plants. However, the induced resistance was not systemic; survival of miners on older, undamaged leaves of damaged seedlings was not different from survival on older leaves of undamaged controls. This localized induced resistance did not produce an overdispersed distribution of either mines or successful mines. On the contrary, the distributions tended towards clumped, although they were not statistically distinguishable from random. Although a localized induced response affected miner survival, no effects on behaviour were observed.     

Effects of leaf and sap feeding insects on photosynthetic rates of goldenrod

Summary Herbivory can alter the balance between sources and sinks within a plant, and changes in the source-sink ratio often lead to changes in plant photosynthetic rates. We investigated how feeding by three insect herbivores affected photosynthetic rates and growth of goldenrod (Solidago altissima). One, a phloem-sap feeding aphid (Uroleucon caligatum), creates an additional sink, and the other two, a leaf-chewing beetle (Trirhabda sp.) and a xylem-sap feeding spittlebug (Philaenus spumarius) both reduce source supply by decreasing leaf area. Plants were grown outside in large pots and insects were placed on them at predetermined densities, with undamaged plants included as controls. All insects were removed after a 12-day feeding period. We measured photosynthetic rates both of damaged leaves and of undamaged leaves that were produced after insect removal. Photosynthetic rates per unit area of damaged leaves were reduced by spittlebug feeding, but not by beetle or aphid feeding. Conductance of spittlebugdamaged leaves did not differ from controls, but internal carbon dioxide concentrations were increased. These results indicate that spittlebug feeding does not cause stomatal closure, but impairs fixation within the leaf. Effects of spittlebug feeding on photosynthetic rates persisted after the insects were removed from the plants. Photosynthetic rates per unit area of leaves produced after insect removal on spittlegug-damaged plants were lower than control levels, even though the measurements were taken 12 days after insect removal. The measurement leaf on spittlebugdamaged plants was reduced in area by 27% relative to the controls, but specific leaf area (leaf area/leaf weight) was increased by 18%. Because of the shift in specific leaf area, photosynthetic rates were also examined per unit leaf weight, and when this was done there were no significant differences between control and spittlebug-damaged plants. Beetle and aphid feeding had no effects on the photosynthetic rate of the leaves produced after insect removal. Plant relative growth rates (in terms of height) were reduced by spittlebugs during the period that the insects were feeding on the plants. Relative growth rates of spittlebug-damaged plants were increased above control levels after insect removal, but these plants were still shorter than controls 17 days after insect removal. Beetles and aphids did not affect plant relative growth rates or plant height. Feeding by both spittlebugs and beetles reduced leaf area, and the effect of the spittlebug was more severe than that of the beetle. These results show that effects of herbivory on photosynthetic rates cannot be predicted simply by considering changes in the source-sink ratio, and that spittlebug feeding is more damaging to the host plant than beetle or aphid feeding.     

Foliar nitrogen and larval parasitism as determinants of leafminer distribution patterns on Spartina alterniflora

ABSTRACT, 1. In northwest Florida, Hydrellia valida hew (Diptera: Ephydridae) frequently mines the leaves of salt marsh cord, grass, Spartina alterniflora . Larvae and pupae are more commonly found in shoreline plants within 3 m of the sea than in inland plants.
2. Shoreline plants contain over twice as much total foliar nitrogen as inland plants. In transects from inland to shoreline plants, leafminer density is positively correlated with total leaf nitrogen content.
3. We experimentally increased foliar nitrogen content on inland Spartina patches to levels above those of shoreline plants. Leafminer densities on experimental centres remained far below those on control edges.
4. Larval parasitism by Opius sp. (Hymenoptera: Braconidae) and Pteromalus sp. (Hymenoptera: Pteromalidae) reaches over 90%, with parasitism increasing from shoreline to inland plants. We suggest that gradients in leafminer densities are caused by differences in parasitoid abundance.     

Effect of mechanical damage on emission of volatile organic compounds from plant leaves and implications for evaluation of host plant specificity of prospective biological control agents of weeds

Assessment of host plant specificity is a critical step in the evaluation of classical biological control agents of weeds which is necessary for avoiding possible damage to non-target plants. Volatile organic compounds (VOCs) emitted by plants likely play an important role in determining which plants attract and are accepted by a prospective arthropod agent. However, current methods to evaluate host plant specificity usually rely on empirical choice and no-choice behavioural experiments, with little knowledge about what chemical or physical attributes are stimulating the insect. We conducted experiments to measure the quantitative and qualitative effects on emission of VOCs caused by simple mechanical damage to leaves of plants known to differ in suitability and attractiveness to a prospective agent. More VOCs were detected from damaged than from undamaged leaves for all three species tested. Discriminant analysis was able to correctly distinguish the taxonomic identity of all plants based on their VOC profiles; however, the VOCs that discriminated species among undamaged leaves were completely different from those that discriminated among damaged leaves. Thus, damaged and undamaged plants present different VOC profiles to insects, which should be considered when conducting host plant specificity experiments. An unacceptable non-target plant, Centaurea cineraria, emitted all except one of the VOCs that were emitted by its preferred host plant, Centaurea solstitialis, indicating the importance of compounds that are repellant in host plant specificity. Centaurea cyanus emitted fewer VOCs than C. solstitialis, which suggests that it lacked some VOCs important for host plant recognition.