170 Ridsdill-Smith et al. Archives of Insect Biochemistry and Physiology 51:170–181 (2002) Induced Responses in Clover to an Herbaceous Mite James Ridsdill-Smith,* Emilio Ghisalberti, and Yong Jiang Halotydeus destructor feeding on subterranean clover cotyledons can cause severe damage. The mites live on the soil surface and move up onto plants to feed. Foraging behaviour consists of palpating, probing, and feeding with frequent transitions between them. Sustained feeding is made up of a series of short (1–2 min) feeds separated by periods of palpating. The mites tend to feed in aggregations, and are attracted to cotyledons damaged by other mites feeding or by mechanical damage. Mites can distinguish between resistant and susceptible cotyledons within 30 min and resistance is antixenotic due to deterrence. Study of the mechanisms shows this to be induced plant resistance. Several green leaf volatiles are involved in the plant/mite interaction. After feeding commences, 2-E-hexenal is released that at low concentrations is attractive to mites, perhaps causing the feeding aggregations. The wound-induced C8 compound, 1-octen-3-one, plays a significant role in the deterrence of cotyledons of resistant subterranean clover varieties to H. destructor. Damaged cotyledons of resistant varieties produce more 1-octen-3-one that those of susceptible varieties. Screening for resistance has identified varieties from Italy showing resistance. H. destructor does not occur in Europe. Production of damage-induced volatiles by the resistant plants may have resulted from invasion by herbivores or pathogens, but not from coevolution with these mites. The responses of H. destructor are probably an adaptation to these general plant defensive compounds. Arch. Insect Biochem. Physiol. 51:170– 181, 2002. © 2002 Wiley-Liss, Inc. KEYWORDS: induced plant resistance; Halotydeus destructor; subterranean clover; 1-octen-3-one INTRODUCTION Redlegged earth mite, Halotydeus destructor (Tucker) (Acari: Penthaleidae), was accidentally introduced to Australia in 1917 from South Africa and spread through southern Australia by 1934 (Swan, 1934; Ridsdill-Smith, 1997; Qin, 1997). Mostly it is found on the soil surface, but moves up onto plants to feed (Gaull and Ridsdill-Smith, 1996). Ridsdill-Smith (1997) reviewed the biology and control of this species. It is a truly polyphagous species with a very wide host plant range, but is considered a particular pest of annual clover plants in pastures, such as subterranean clover, Trifolium subterraneum (L.). Redlegged earth mite is active for about 6 months of the year during the cool moist winter months, spending the summer as a diapausing egg (Ridsdill-Smith and Annells, 1997). Populations of 12,000 mites/m2 are common in pastures of southern Australia and cause severe losses of productivity to the pastures. Mites feed on all stages of clover in pastures, but the greatest damage occurs in autumn when oversummering mites hatch at much the same times as the annual clover germinates. At these times, high mite populations can destroy clover seedlings. Thus, a high priority for the clover- breeding program is to develop resistance in subterraneum clover cotyledons to protect the seedlings. Screening methods to test for resistance in clover seedlings against H. destructor were developed by Gillespie (1993). The screening is carried out in the glasshouse with rows of different varieties in wooden (or plastic) containers at a rate of about seven varieties per container, and the mites are thus given a choice. Feeding damage is assessed after 1 Centre for Legumes in Mediterranean Agriculture, University of Western Australia, Australia Contract grant sponsor: Australian Wool Innovation. *Correspondence to: Dr. James Ridsdill-Smith, CSIRO Entomology, Private Bag No 5, Wembley, WA 6913, Australia. E-mail: James.Ridsdill-Smith@csiro.au Received 6 May 2002; Accepted 22 August 2002 © 2002 Wiley-Liss, Inc. DOI: 10.1002/arch.10063 Published online in Wiley InterScience (www.interscience.wiley.com) Archives of Insect Biochemistry and Physiology Induced Resistance to Herbaceous Mite and 2 weeks. Levels of resistance to cotyledons are assessed using a visual scale of 1–10 where 1 is no feeding damage, through to 5 where half the upper cotyledon surface has feeding damage (silvering patch) to 10 where the whole cotyledon is dead or dying. Of about 7,500 subterranean clover introductions to Australia, 675 have been screened with mites and 18 show seedling resistance (Ridsdill-Smith and Nichols, 1998). Thirteen of the resistant introductions were collected in Italy, of which 10 were from Sicily. The redlegged earth mite does not occur in Europe at all, and so this resistance is not the result of coevolution. It is likely that this seedling resistance in subterranean clover is due to general plant defense mechanisms that protect the plants against other biotic stresses, and that the mites have adapted to respond to them. Over a number of years, different aspects of the plant-insect interactions involved in T. subterraneum cotyledon resistance to H. destructor have been investigated. A picture has developed of an induced resistance mechanism involving the production of plant volatiles that deter mites from feeding, but that are not specific responses to mite feeding. The term “induced resistance” is used in the present context to refer to a rapid response by the plant to biotic or abiotic challenges. In this study, we review the current state of knowledge of the mechanisms involved in the resistance, and consider how this reflects on our understanding of plant defense systems. MITE FEEDING BEHAVIOUR In pastures, most H. destructor are on or near the soil surface with only 10% at any one time on the upper canopy (Gaull and Ridsdill-Smith, 1996). The mites move up onto the canopy of the plants where they feed on the upper adaxial surfaces of the leaves. In the annual pastures of southwestern Australia, the mites show a preference for feeding on subterranean clover leaves. While the proportion of mites is distributed approximately in proportion to the relative abundance of the subterranean clover in the pasture (50%), the proportion feeding was higher on the December 2002 171 clover (74%) (Gaull and Ridsdill-Smith, 1996). This indicates that the mites are selecting their hosts for feeding after searching on all the plants in the pasture. A characteristic of the feeding mites is that they are in aggregations of 5–6 mites. In the field, an average of 88% of aggregated individuals are feeding, whereas only 23% of solitary individuals are feeding (Gaull and Ridsdill-Smith, 1996). The mites feeding in aggregations leave patches of feeding damage, and this provides an easy way to visually assess the amount of feeding from the size of the patch. The aggregations provide a benefit to the feeding mites, in terms of increased mean weight gain over a 2-h period (Gaull and Ridsdill-Smith, 1997). Feeding mites require applying mechanical physical force to penetrate the upper epidermis of the plants (Jiang and RidsdillSmith, 1996b), and it is possible that the aggregations may assist mites in penetrating the epidermis. When feeding, the mite braces its legs on the epidermis by clamping the tarsal claws into the surface, leaving triangular dents (Ridsdill-Smith et al., 1997). One hole 3 mm in diameter is made in each cell with the movable digit of the chelicera, and the mites suck out the cell contents with a pharyngeal pump. The side-walls of the cells collapse, and the upper surface changes from convex to flat. Air in the cells probably causes the silvery appearance. Feeding is restricted to the epidermal and subepidermal cells, which is where resistance factors are likely to be found. The repertoire of foraging behaviour has three elements. Palpating is where the mites run on the leaf surface continually tapping it with the first pair of legs and palps. Movement is continual and directional. Probing is where there are regular stops, continual changing of position, and momentarily the mouthparts are applied to the plant surface. Feeding is where the mite is stationary, with legs braced and mouthparts applied to the leaf surface. Also, occasionally the mite is stationary and immobile with the first pair of legs held out straight in front of the mite. Host acceptance behaviour has been observed using starved marked mites in a group. During a 70-min observation period, mites spend 66% of their time in aggregations, and 83% 172 Ridsdill-Smith et al. of that is feeding, while solo/pairs of mites only spend 65% of time feeding (Gaull and RidsdillSmith, 1997). RESPONSES TO SEEDLINGS OF SUSCEPTIBLE AND RESISTANT VARIETIES the two cotyledons divided by the total number of mites on both cotyledons. The number of mites on cotyledons is positively correlated with feeding damage over 3 h (67% of variability), and damage scores of cotyledons from 1–2 week screening choice experiments are correlated with deterrence on the same varieties in 3-h choice experiments (93% of variability). The choice made by mites in the 3-h experiments is relative. Mites show 74% deterrence to resistant variety 1 compared to a susceptible control, and 45% to resistant variety 2 against the same control, but they also show 83% deterrence to resistant variety 2 compared with resistant variety 1 (Fig. 1). While the total time spent feeding by a starved marked mite is significantly greater on the cotyledon of a susceptible than a resistant variety, the mean duration of each separate feeding event is similar on both at around 1.5–2 min (Gaull and Ridsdill-Smith, 1997). The mites are continually moving between probing and feeding. The transitions from probing to feeding and back on the resistant variety are 62% of those on the susceptible variety. The mites have only rudimentary eyes and chemical cues are likely to be important in host plant acceptance. When H. destructor are confined over 7 or 14 days to varieties of subterranean clover that are resistant and susceptible with a single variety per pot in a non-choice situation, mites on resistant varieties produce 45% of the progeny of mites on susceptible varieties, and feeding damage on resistant varieties is 45% of feeding damage on susceptible varieties (Ridsdill-Smith, 1995). When mites are given a choice of seedlings of a resistant and a susceptible variety in the same pot over 7 days, feeding damage on the resistant varieties is only 23% of that on the susceptible variety. Mites also readily discriminate between detached cotyledons of resistant and susceptible varieties in a 3-h choice tests on soil in a Petri dish (Jiang and Ridsdill-Smith, 1996a). When mites are given a choice of two detached cotyledons of the same variety, similar numbers occur on each, but given a choice between a resistant and a susceptible variety, the choice is strong and occurs within 30 min. Thus, the mites are distinguishing the resistant cotyledons by antixenosis or deterrence. This deterrence is measured by the difference in numbers of mites on Cotyledons damaged by H. destructor feeding are more attractive to mites in choice trials than un- Fig. 1. Mite preference in choice between subterranean clover cotyledons over 3 h. Mites showed a clear choice between resistant and susceptible varieties, but were also able to distinguish between the two resistant varieties. Data from Jiang and Ridsdill-Smith (1996a) and Wang et al. (2001). RESPONSES TO DAMAGED COTYLEDONS Archives of Insect Biochemistry and Physiology Induced Resistance to Herbaceous Mite damaged cotyledons (Jiang and Ridsdill-Smith, 1996a). Individual marked mites spend twice as much time feeding and twice as long in total activities on cotyledons previously damaged by mite feeding in a choice situation than on a previously undamaged cotyledon (Gaull and Ridsdill-Smith, 1997). When mechanical cuts (about 15) are made with a razor blade in the upper surface of the cotyledons, the exposure of the cell contents also increased the attractiveness of the cotyledon relative to an undamaged control (Jiang and Ridsdill-Smith, 1996a) (Fig. 2). However, while mechanical damage increases the attractiveness of cotyledons both of a susceptible and a resistant variety, there are still more mites on the damaged susceptible than the damaged resistant cotyledon (Jiang and Ridsdill-Smith, 1996a). Damaged cotyledons remain attractive to mites for some time after the damage occurs when the damage is caused by mite feeding or mechanically (Jiang et al., 1997). For the mite feeding damage treatment, 200 mites are released into pots containing 15 seedlings and allowed to feed for 18 h, before being removed with a compressed air sucker. For the mechanical damage treatment, a spiral metal screw is pressed into the leaf in 10 places. Damaged and undamaged plants are grown in separate pots, and detached when ready to test at 1.5 h (feeding damage), 3 h (mechanical damage), 1, 3, and 7 days after damage had occurred. Mites are given a choice of a damaged or an undamaged 173 cotyledon, and numbers of mites on each cotyledon counted over 3 h. Preference is expressed from the average number per cotyledon. Mites prefer damaged cotyledons for 4 days on the susceptible variety, and 7 days on the resistant variety (Fig. 3). Preference is greater with mite damaged than mechanically damaged cotyledons on most occasions, especially for the resistant variety. The plants are growing in soil and presumably continue to produce the factor that attracts mites from damaged cotyledons. EFFECTS OF VOLATILES FROM THE PLANTS Since we were unable to establish that secondary metabolites were involved in the resistance mechanism in subterranean clover cotyledons, we tested the possible role of volatile compounds. The method used successfully by other authors on mites was a glass tube olfactometer (Dabrowski and Rodriguez, 1971). However, for H. destructor the response to volatile compounds depends on the method of their presentation to the mites (Jiang et al., 1996b). In a glass tube assay, two flasks (8 cm diameter) joined by a U-shaped glass tube (2 cm diameter) is used with screen cloths across the tube at the joints with the flasks to limit mite movement. Cotyledons from a susceptible subterranean clover variety are frozen in liquid nitrogen and ground up in a mortar. Different weights of the powder are placed in one of the flasks, and Fig. 2. Mite preference in choice between an undamaged control and a mechanically damaged subterranean clover cotyledon over 3 h. Mites show a preference for mechanically damaged cotyledons both with a susceptible and a resistant subterranean clover variety. Data from Jiang and Ridsdill-Smith (1996a). December 2002 174 Ridsdill-Smith et al. Fig. 3. Mite preference in choice between subterranean clover cotyledons with mite feeding damage or mechanical damage over time for 7 days. Each bioassay lasted for 3 h with mites given a choice between freshly detached cotyledons. Feeding damage was the result of feeding by 200 mites in pots containing 15 seedlings of resistant or susceptible varieties. Mechanical damage was the result of pressing the tip of a screw into the upper surface of a cotyledon in 10 places. Plants were then left attached to the growing seedling until testing. Mites detected damaged cotyledons for up to 7 days. Data from Jiang et al. (1997). the other flask contains no powder as a control. Forty mites are released in the middle and counted after 5 and 10 min. There is no consistent difference in mite numbers in relation to the amounts of tissue, and the mites do not seem to be able to detect the directions from which the volatiles are coming (Fig. 4). In the second method, volatiles are presented in a Parafilm membrane sachet with 1% glucose added as a phagostimulant. The relative concentrations of volatiles are presented at 4, 40, 100, and 400% of the in vivo concentration found in a susceptible variety (labelled 1 to 4 in Fig. 4). Mites respond strongly to concentrations when presented in this system, being attracted at concentrations below 100% in vivo and deterred at 400% in vivo (Fig. 4). The membrane sachet test provides the better test to detect concentration effects of volatiles. It is clear that the method of presentation is important for detecting effects of volatile compounds on H. destructor behaviour through olfactory and gustatory responses. Volatile compounds produced from damaged subterranean clover cotyledons are trapped onto charcoal and identified using GC/MS. The major compounds collected are similar from susceptible and resistant varieties but differ in their relative amounts. The major compounds were 1-octen-3one, 1-octen-3-ol, and 2-E-hexenal, making up 77% of the volatile fraction (Jiang et al., 1997; Wang et al., 2001). Levels of these volatile compounds are determined in the headspace of artificially damaged cotyledons of six resistant and four susceptible subterranean clover varieties and expressed relative to n-hexanol (Jiang et al., 1996a). Feeding damage scores for cotyledons on seedlings of these varieties are measured in separate screening trials over 2 weeks in choice situations (Gillespie, personal communication). Quadratic regressions of feeding damage as a function of concentration (log scale) are used to determine which compounds best predict damage. The r2 is greater for 1-octen3-one (0.834) than for 1-octen-3-ol (0.307) or 2E-hexenal (0.321), suggesting a key role for 1octen-3-one in the cotyledon resistance. Archives of Insect Biochemistry and Physiology Induced Resistance to Herbaceous Mite 175 ROLE OF 1-OCTEN-3-ONE Fig. 4. Mite preference for volatiles from a susceptible subterranean clover variety and presented in two different bioassays. In the glass tube bioassay, mites had a choice between different quantities of ground-up cotyledon tissue (0.04, 0.10, 0.50, or 5 g) and air. The distribution of mites was measured after 5 and 10 min. In the Parafilm sachet, mites had a choice of volatiles collected from the headspace of crushed cotyledons, and the control was without the volatiles. The volatiles were dissolved in Tween 80 (a detergent) to which 1% glucose was added, and made up to 4, 40, 100, or 400% of in vivo concentration. Mite numbers on each membrane were counted over 3 h. For each method, concentrations are expressed as 1 to 4, although the concentrations the mites were exposed to may not have been completely comparable. Mites responded better to the Parafilm sachet bioassay. Data from Jiang et al. (1996b). Fig. 5. Mite preference in 3-h bioassays for cotyledons of a susceptible and a resistant subterranean clover variety grown in the glasshouse with and without shade. Shade provided by placing a cardboard box over the seedlings with four openings allowing dim light. Cotyledons were December 2002 In a separate experiment, cotyledons of a susceptible variety and a resistant variety, grown under shading, are completely avoided by mites in choice feeding tests (Jiang et al., 1996a) (Fig. 5). Analysis of the volatile compounds show that the levels of the C8 volatile compounds increase by more than 50%, whereas 2-E-hexenal decreases by 9% in both shaded varieties. These results emphasize the significance of 1-octen-3-one (1) in resistance (Jiang et al., 1996a). The influence of the volatile compounds 1octen-3-one and 2-E-hexenal on H. destructor was investigated by placing droplets on the surface of a cotyledon of a susceptible subterranean clover variety (Jiang et al., 1997). The preference or deterrence of mites was measured in choice tests using cotyledons with droplets containing the compound, a solubilizing agent and 1% glucose, and the control with no compound. The treatments consist of a range of concentrations, applied either as a single 5-ml droplet on the basal part of the cotyledon (where mites do not usually feed), or as five 0.5-ml droplets spread over the cotyledon surface. The volatile fraction from cotyledons was tested at levels from 4 to 400% of in vivo, used 12–13 days after sowing. Concentrations of octenone were measured from shaded and unshaded seedlings. Mites were adversely affected by the octenone in the shaded cotyledons. Data from Jiang et al. (1996a). 176 Ridsdill-Smith et al. and the individual metabolites at levels equal to or above the in vivo level. At low concentrations, mites show a significant preference for both compounds, and at high concentrations a significant deterrence (Fig. 6). Mites are attracted to cotyledons with five scattered droplets of 2-E-hexenal at concentrations 50–1,000 ppm, while with 1octen-3-one the cotyledon with droplets is attractive at 1 ppm and deterrent at 10,000 ppm. In contrast, with a single larger droplet there is a much stronger concentration-dependent effect (Fig. 6). 2-E-hexenal is attractive at 10–100 ppm and deterrent at 10,000 ppm, while 1-octen-3-one is not attractive, but deters mites at 100 ppm and all mites are killed within 3 h at 10,000 ppm. From these results, we conclude that 2-E-hexenal is probably the factor that attracts mites to damaged cotyledons, including those on which mites are feeding, and 1-octen-3-one is the volatile compound that is a primary cause of the antixenotic deterrence of mites from cotyledons of resistant subterranean clover varieties. A contribution of the other minor volatile compounds to resistance cannot be ruled out. Fig. 6. Mite preferences in 3-h bioassays for cotyledons of susceptible subterranean clover variety droplets containing different concentrations of 2-E- hexenal or 1-octen-3one. There was one 5-ml droplet per cotyledon or five 0.05-ml droplets, with controls of cotyledons with droplets without the compounds. Preference decreased with INDUCED RESISTANCE: DISCUSSION AND CONCLUSIONS H. destructor spend 90% of their time on or near the soil. They move up to feed in the upper canopy of pastures, on the dorsal (adaxial) surface of leaves. Most feeding occurs in aggregations, and mites in such groups benefit from greater weight gain. The mites are polyphagous with a very high range of higher plants on which they can feed (Ridsdill-Smith, 1997). These range from species on which populations increase rapidly, to plants where mites cause damage but survival is low and few progeny are produced. In addition, they are able to maintain populations feeding on lower plants at the soil surface (McDonald et al., 1995). Mite foraging behaviour is a simple sequence from palpating to probing to feeding, but sustained feeding is made up of a series of short (1–2 min) bouts of feeding followed by a return to probing. In this way, the mite is continually sampling its environment, and is able to detect rapidly between plants for feeding. This ability to choose also provides for us a useful tool for studying the mechanisms of resistance, but also means that mite choices are increasing concentration, mites being attracted at low concentrations and deterred at high concentrations. Octenone was more deterrent than hexenal, and concentration-dependent responses were greater with the larger droplet. Data from Jiang et al. (1997). Archives of Insect Biochemistry and Physiology Induced Resistance to Herbaceous Mite made using relative differences. For example, the mites do select between two resistant subterranean clover varieties. In subterranean clover cotyledons, the green leaf compounds identified as affecting H. destructor, act as attractants at low concentrations and deterrents at high concentrations. The mites respond in a dose-dependent way when the volatiles are presented in a Parafilm membrane with glucose as a feeding stimulant. This mimics the way that volatiles would be affecting mites on a plant. The response is initiated on plants after the initial probing and feeding, and thus is induced. Two volatile compounds, 2-E-hexenal and 1-octen-3-one, are mainly involved. It is the quantity produced that affects mite behaviour. Thus 2-E-hexenal in general seems to be acting to attract mites, which may help explain how feeding aggregations are formed. However, 1-octen-3-one seems to be the cause of deterrence. Mites are more attracted to damaged than undamaged cotyledons of a resistant variety. Presumably, the concentration of volatiles decreases with distance from the point of feeding puncture, and may thus attract mites from a distance (on the cotyledon), but when they arrive at the feeding site concentrations of 1-octen-3-one may be high enough to cause deterrence. The induced responses occur whether the damage is from feeding mites or artificial/mechanical and so are not specific to the mites. Indeed, natural resistance is found in subterranean clover varieties collected in Italy where these mites do not occur, and so is not coevolved. It has been shown that fatty acid oxidation products from the lipoxygenase pathway can play an important part in plant defenses against attack of herbivorous insects or animals, fungal or bacterial pathogens. Recent studies indicate that the response of plants to different types of wounding involve common responses as well as specific ones determined by the cause of the damage (Pickett and Poppy, 2001). Considerable interest has been shown in the observations that the production of volatile compounds at the wound site increases significantly. The volatiles produced in general wound responses arise from the catalytic activity December 2002 177 of lypoxygenases (LOX), a group of enzymes believed to be present in all green plants (Gardner, 1991). These enzymes operate, principally, on the two important fatty acids linoleic and linolenic acid and produce hydroperoxides that are degraded by hydroperoxy lyases (HPL) to a suite of C6-aldehyde and alcohol compounds (Hatanaka, 1993; Mau et al., 1994) (Fig. 7). These compounds appear to possess signalling properties. They are released in sufficient quantities to be detected by animals, can trigger formation of phytoalexins, and reduce insect feeding rates. Importantly, they have been shown to induce a subset of defenserelated genes such as those involved in the phenylpropanoid pathway and the LOX pathway (Bate and Rothstein, 1998). In the response of T. subterraneum to leaf wounding, the formation of 2-E-hexenal is expected. However, the production of 1-octen-3-one and 1-octen-3-ol is unusual; only some species of fungi are known to produce them in significant quantities (Mau et al., 1994). Moreover, 2-Ehexenal is known to be produced from linolenic acid but 1-octen-3-one and 1-octen-3-ol arise from linoleic acid (only the R-enantiomer of 1-octen-3ol is shown in Figure 8, but the formation of the S-enantiomer is not excluded). The biosynthesis of 1-octen-3-ol from linoleic acid in the mushroom Agaricus bisporus has been studied (Mau et al., 1994; Gardner, 1991). It was found that the 10-hydroperoxy derivative was the product of lipoxygenase action and the precursor of 1-octen-3-ol and 1octen-3-one (Fig. 8). The observation that resistant varieties of T. subterraneum produced more 1-octen3-one than susceptible varieties, although not proven, suggests that this difference reflects different amounts of the precursor fatty acids in the cotyledons. In the experiment described here, where plants grown in shade deter mites, it is the increase in 1-octen-3-one that causes the deterrence. The greater production of the C8-volatiles in dark-grown leaves is a consequence of the higher levels of the precursor linoleic acid produced under this condition (Harwood, 1980). The C6-volatiles induce some defense genes, but at a lower level and a narrower range than 178 Ridsdill-Smith et al. Fig. 7. Basic reactions of the lipoxygenase pathway for the formation of some C6-volatile compounds from linoleic acid (A) and linolenic acid (B). those induced by jasmonic acid and methyl jasmonate, which are also LOX-derived metabolites (Bate and Rothstein, 1998). In contrast, the gene-inducing ability of the C8-compounds produced by T. subterraneum has not been studied. The only attempt to determine if C8 compounds induced defense-related genes was a test involving oct-2-en-1-ol, which was found to have no effect on the induction of LOX mRNA (Bate and Rothstein, 1998). Spider mites, Tetranychus species, avoid cotton seedlings on which co-specifics have fed (Karban and Carey, 1984; Harrison and Karban, 1986). A range of compounds is produced when spider mites feed on lima bean and cucumber leaves and the blend of compounds varies between the plants (Dicke et al., 1990). Of these compounds, terpenes and a phenol are attractive to predatory mites, and Dicke et al. (1990) argue that they are used by predators as cues to find their herbivorous prey. These compounds are produced by spider mite feeding but not by artificial damage, in contrast to the situation with H. destructor. Terpenoids do not appear to be present in detectable amounts in damArchives of Insect Biochemistry and Physiology Induced Resistance to Herbaceous Mite Fig. 8. Steps of the lipoxygenase pathway for the formation of C8-volatile compounds from linoleic acid. aged cotyledons of subterranean clover (Jiang et al., 1996a). The resistance mechanisms described here are one of several that affect H. destructor feeding on different plants. In trifoliate leaves of subterranean clover, resistance is due to isoflavonoids (Wang et al., 1998), in Trifolium glanduliferum, the volatile compounds coumarin and b-ionone (Wang et al., 1999), while in yellow lupins the resistance mechanism is alkaloids (Wang et al., 2000). However, the mechanism in subterranean clover cotyledons is the only one that appears to be induced by mite feeding. ACKNOWLEDGMENTS The authors are grateful to Jenny Reidy Crofts who carried out most of the bioassays, Tanya Pican and Celia Pavri who helped with many of the experiments, and Dennis Gillespie and Phil Nichols for advice on screening for resistance and supplying the seed. Dr. Shao Fang Wang provided much valuable discussion on resistance mechanisms. The work has been supported by Australian wool growers through the Australian Wool Innovation. December 2002 179 LITERATURE CITED Bate NJ, Rothstein SJ. 1998. C6-volatiles derived from the lipoxygenase pathway induce a subset of defense-related genes. Plant J 16:561–569. Dabrowski ZT, Rodriguez JG, 1971. Studies on resistance of strawberries to mites. 3. Preference and nonpreference responses of Tetranychus urticae and T. turkestani to essential oils of foliage. J Econ Ent 64:387–391. Dicke M, Sabelis M, Takabayashi J, Bruin J, Posthumus MA. 1990. Plant strategies of manipulating predator-prey interactions through allelochemicals: prospect for application in pest control. J Chem Ecol 16:3091–3118. Gardner HW. 1991. Recent investigations into the lipoxygenase pathway of plants. Biochim Biophys Acta 1084:221–239. Gaull KR, Ridsdill-Smith TJ. 1996. The foraging behaviour of redlegged earth mite, Halotydeus destructor (Acarina: Penthaleidae), in an annual subterranean clover pasture. Bull Ent Res 86:247–252. Gaull KR, Ridsdill-Smith TJ. 1997. Host acceptance behaviour in Halotydeus destructor (Penthaleidae: Acarina). J Ins Beh 10:859–869. 180 Ridsdill-Smith et al. Gillespie DJ. 1993. Redlegged earth mite (Halotydeus destructor) resistance in annual legumes. In: Delfosse E, editor. Pests of pastures. CSIRO: Melbourne. p 211–213. Harrison S, Karban R. 1986. Behavioural response of spider mites (Tetranychus urticae) to induced resistance of cotton plants. Ecol Entomol 11:181–188. Harwood LJ, 1980. Plant acyl lipids: structure, distribution, and analysis. In: Stumpf PK, editor. The biochemistry of plants, a comprehensive treatise. vol. 4. Lipids: structure and function. New York: Academic Press. p 1–55. editor. Spices, herbs and edible fungi: Developments in food science 34. Elsevier: Amsterdam. p 657–684. Pickett J A, Poppy G M. 2001. Switching on plant genes by external chemical signals. Trends Plant Sci 6:137–139. Qin TK. 1997. Population genetics of the redlegged earth mites, Halotydeus destructor and H. anthropus (Acarina: Penthaleidae) from Australia and/or South Africa. Bull Ent Res 87:289–298. Hatanaka A. 1993. The biogeneration of green odour by green leaves. Phytochemistry 34:1201–1218 Ridsdill-Smith TJ. 1995. Responses and feeding damage of redlegged earth mite (Acarina: Penthaleidae) to seedlings of resistant and susceptible subterranean clover varieties. Aust J Agric Res 46:1091–1099. Jiang Y, Ridsdill-Smith TJ. 1996a. Antixenotic resistance of subterranean clover cotyledons to redlegged earth mite, Halotydeus destructor. Entomol Exp Appl 79:161–169 Ridsdill-Smith TJ. 1997. Biology and control of Halotydeus destructor (Penthaleidae: Acarina): a review. Exp Appl Acarol 21:195–224. Jiang Y, Ridsdill-Smith TJ. 1996b. Examination of the involvement of mechanical strength in antixenotic resistance of subterranean clover cotyledons to the redlegged earth mite, Halotydeus destructor (Acarina: Penthaleidae). Bull Ent Res 86:263–270. Ridsdill-Smith TJ, Annells AJ. 1997. Seasonal occurrence and abundance of redlegged earth mite Halotydeus destructor (Acari: Penthaleidae) in annual pastures of southwestern Australia. Bull Ent Res 87:413–423. Jiang Y, Ghisalberti E, Ridsdill-Smith TJ. 1996a. Correlation of 1-octen-3-one with antixenotic resistance in subterranean clover cotyledons to redlegged earth mite, Halotydeus destructor (Acarina: Penthaleidae). J Chem Ecol 22: 369–382. Ridsdill-Smith TJ, Nichols PGH. 1998. Development of pasture legumes resistant to redlegged earth mite. In: Zalucki MP, Drew RAI, White GG, editors. Proc. 6th Australasian Appl Ent Res Conf, Brisbane. University of Queensland, Brisbane. p 382–389. Jiang Y, Ridsdill-Smith TJ, Ghisalberti E. 1996b. Assays for the effects of volatile compounds from damaged cotyledons of subterranean clover on the redlegged earth mite, Halotydeus destructor. Exp Appl Acarol 20:61–72. Ridsdill-Smith TJ, Craig S, Beaton CD. 1997. Microscopic examination of feeding damage to subterranean clover cotyledons caused by Halotydeus destructor (Tucker) (Acari: Penthaleidae). In: Mitchell R, Horn DJ, Needham GR, Wellbourn WC, editors. Proc. IX Int. Congress Acarology. Ohio Biological Survey: Columbus, Ohio. p 55–57. Jiang Y, Ridsdill-Smith TJ, Ghisalberti EL. 1997. The effect of volatile metabolites of lipid peroxidation on the aggregation of redlegged earth mites Halotydeus destructor (Acarina: Penthaleidae) on damaged cotyledons of subterranean clover. J Chem Ecol 23:163–174. Swan DC. 1934. The red-legged earth mite Halotydeus destructor (Tucker) in South Australia: with remarks upon Penthaleus major (Dugès). J Dept Agr S Aust 38:353–367. Karban R, Carey JR. 1984. Induced resistance of cotton seedlings to mites. Science 225:53–54. Wang SF, Ridsdill-Smith TJ, Ghisalberti EL 1998. Role of isoflavonoids in resistance of subterranean clover trifoliates to the redlegged earth mite Halotydeus destructor. J Chem Ecol 24:2089–2100. McDonald G, Moritz K, Merton E, Hoffmann AA. 1995. The biology and behaviour of redlegged earth mite and blue oat mite on crop plants. Plant Prot Q 10:52–54. Mau J, Beelman RB, Ziegler GR. 1994. Aroma and flavor components of cultivated mushrooms. In: Charalambous G, Wang SF, Ghisalberti EL, Ridsdill-Smith TJ. 1999. Volatiles from Trifolium as feeding deterrents of redlegged earth mites. Phytochemistry 52:601–605. Wang SF, Liu AY, Ridsdill-Smith TJ, Ghisalberti EL. 2000. Archives of Insect Biochemistry and Physiology Induced Resistance to Herbaceous Mite Role of alkaloids in resistance of yellow lupin to redlegged earth mite Halotydeus destructor. J Chem Ecol 26: 429–441. Wang SF, Ridsdill-Smith TJ, Ghisalberti EL. 2001. Resistance December 2002 181 in pasture legumes to redlegged earth mite Halotydeus destructor (Tucker) (Acari: Penthaleidae). In: Halliday RB, Walter DE, Proctor HC, Norton RA, Colloff MJ, editors. 10th Int Congr Acarology. Melbourne: CSIRO. p 328–341.