INVESTIGATIONS ON TOXIC, ANTIFEEDANT AND REPELLENT PROPERTIES OF THE DEFENSIVE SECRETION OF CORIDIUS JANUS(HEMIPTERA;PENTATOMIDAE) AND A SYNTHETIC MIXTURE OF ITS MAJOR VOLATILE CONSTITUENTS

Effectiveness of defensive secretion of Lbridius janrcs and a synthetic mixture of its two major volatiles, t-2-hexenal and n-tridecane (60:40 wlw) was assayed as topical applicants, fumigants, antifeedants, irritants and repellents against neven insect pesta, Periplaneta amcricana, Sitophilus oryzac, Tribolium castancum, .Musea damestica, Oecophylla smaragdina, Eurcma hecabe and Oysdcrcus c i n g u b . The effectiveneu of the test materials was found to be dependent on their mode of application and the species of test insect. There was a parallelism in the activities of the natural defensive e r e t i o n and, the synthetic mixture. Generally, both test materials p e r f o d poorly when tested as topical applicants and fumigants, although, the nah~ral defensive secretion showed some effect as a topical applicant against adults of 5. o w e , and M. damcstica (LD, < kg). N t h test materials were remarkably effective antifeedants against S. o p e and E. hccabc. These also proved equipotent as irritants against P. amcricana and as repellents to D. cingulanLF and 0. smaragdina. It is suggested that the synthetic chemical mixture may be used for effective insect pest control.


Introduction
Natural pesticides of plant origin such as pyrethrins and cinerins have long been used in the control of various insect pests. In the recent past, secretions of insect origin have also received attention. Consequently, insect pheromones, insect juvenile hormone and its analogues have been recognized and are presently being used as a part of the integrated pest management strategies with varying degrees of success. In contrast, a large number of insect defensive secretions have been characterised 2> 47 l7 although no effort has been made in using these as possible agents for pest control strategies. Nevertheless several biological aspects of insect defensive secretions such as toxic proper tie^,^ alarm pheromone activity and repellent properties have been recognized in insect defensive secretions.
Coridius janus Fabricius (Hemiptera:Pentatomidae) inhabits the dry zone of Sri Lanka. It feeds mainly on sweet pumpkin (Cucurbitaceae) and does not cause a great threat to the host plant unless pest population rises very high. This insect ejects a defensive secretion which has a pungent smell. A previous study1' on the defensive secretion of C. janus has shown that it possesses fairly high fumigant properties towards three test insects viz. Anoplolepis longipes (Formicidae), Culeu quinquefasciatus (Culicidae) and Sitotroga cerealella (Gelechiidae). This study also revealed that the two major volatiles of this secretion, t-Zhexenal and n-tridecane which comprise 90% of the total volatiles of the secretion are mainly responsible for the above fumigant property. Further, a synthetic combination of t-2-hexenal : n-tridecane ( 6 0 , wtw) was shown to possess the highest fumigant activity compared to all other possible combinations of the above chemicals.
The present study was undertaken to compare the effectiveness of the natural defensive secretion with that of a synthetic mixture of its major volatiles t-2-hexena4n-tridecane (60:40, whv), as topical applicants, fimigants, antifeedants, irritants and repellents against a variety of insect orders. This information could be of value for insect pest control strategies using natural products.

Rearing of Con'dius janus and Species of Test Insects
Live individuals of C. janus were collected from their natural habitats, the sweet pumpkin fields at Aulegama, Kurunegala District. In the laboratory, they were caged in small groups and maintained for weeks on fresh leaves of sweet pumpkin.
In this study, seven species of insects belonging to six diierent orders were used as test insects. These insects were collected from their natural habitats and allowed to multiply under laboratory conditions in separate culture containers. The species of test insects used are listed in Table 1, together with the food on which they were maintained in the laboratory. Rearing of four species of test insects., P. americana, M. domestics, S. o p e , T. castaneum in the laboratory were carried out following the methods described by peterson.13 Rearing of 0.srnaragdina in the laboratory was unsuccessful. The insects (soldiers) collected from their natural habitats were, therefore used for the experiments. They could be maintained on glucose and water for 4-6 days until they were used in the experiments.
The adults of E. hecabe were reared on 4% sucrose solution or nectar of flowers in rectangular glass tanks [I' x 2' x 2', 5 male: 5 femaleltank] tightly covered with a piece of mosquito net for ventilation. Two polythene bag pots containing young plants of Kathurumurunga (Sesbania grandiflora) were placed in each tank so that the adult female butterflies could lay eggs on these leaves. When the eggs were hatched, the caterpillars were transferred into plastic bottles (4 L) covered with pieces of mosquito net for ventilation. The caterpillars were reared on leaves of Kathurumurunga. The 2nd instar larvae (1-1.5 cm in length) were used in the experiments.
Nymphs and adults of D. cingulatus were reared on fruits or leaves of Okra in insect cages fitted with metallic mesh. Only the adult stage was used in the experiments.
All insect rearing cages were cleaned regularly. The moisture content of the rearing cages was maintained by keeping moist cotton balls in s m d containers in the cages with the insects. The temperature of the laboratory during the experimental period was 28 + ZOC. UK and found to be 99.9% pure by GLC. Liquid paraffim and acetone were purchased $om BDH, UK.

tion of Test I b l a W s for Bioassay
Naturd defensive secretion of C. janus was obtained by placing insects in preweighed glass vials md pinching their Pegs with fine forceps mtil the defensive secretion was ejected. The insects were removed from the vials immediately after the ejection. The vials were weighed and frozen until they were used in the experiments.
Synthetic mixture , t-2-hexenal/n-tridecane (60:$0,w/w) was prepared by mixing thorsug8ily the neat liquids of 2-hexend and n-tridecanc in the ratio of 6090 (w/w) until a single phase was obtained and weighing out from this mixture appropriate mourllts into s m d glass Gds.

2.4
Stiaclies on Activity of the Defensive Secmtion

Topical applicotr'on
Preliminary studies showed that the fresh defensive secretion of C janus was lethal to Qhe species of insects tested. Immediate death of some species of insects was observed after applykg very small amounts (0,2-0.4 p1) of the defensive secretion onto the thorax of the insects using a micro-applicator. The naturd secretion was therefore diluted with a suitable non toxic solvent. For each test species the suitable non toxic mivent was selected by applying a series of doses (in the range of 0. For studies of activity, the natural defensive secretion was diluted with the selected suitable solvent and various doses (as drops) were applied onto the thorax of each test insect. Insects treated with only the solvent served as controls. After the treatment, the insects were placed separately in recovery cages. Glass vials, tubes or flasks of various sizes (depending on the size of the species of test insect) covered with pieces of mosquito net or perforated plastic caps were used as recovery cages. Mortality was recorded after 24 hours. Each dose was tested three times with different batches of insects and e u m number of 10 insects were used for each dose.
values were calculated by Probit analysis of data as described by p us vine! The activity of the synthetic mixture was also tested as described above.

Fumigantproperty
C. janus in a preweighed small vial was made to eject its defensive secretion. Immediately after the secretion was ejected, C. janus was removed and the vial was closed and weighed. This vial was introduced into an airtight, large, glass container (10L) where test insects (20-30) had been placed about V2 hr prior to the experiments. Small vials containing the defensive secretion or the synthetic mixture were then opened and the volatiles allowed to diffuse into the large glass container. Mortalities were recorded after 24 hours. Small vials were weighed again in order to make sure that all materials had volatilized into the glass container. Distilled water was used in the control experiments. A series of concentrations were tested, and triplicates were run for each test. The LC50 values were calculated by Probit analysis of data.6

Antifeedant activity
Two species of test insects S. olyzae, and E. hecabe were used in this study.

2.4.3e Antifeedant studies using S oryure
The test insects were starved for 24 hours prior to the experiments. Known quantities of rice grains (500 mg/test) were coated evenly (0.2 pVgrain) with acetone, a series of concentrations of the natural secretions, the synthetic mixture diluted with acetone respectively. The treated rice grains were placed separately in labelled 12.5 ml glass vials (500 mg rice/vial). Twenty starved insects were added to each vial and the vials were closed with perforated plastic caps for ventilation. A group of insects in a vial with untreated rice grains was used as control. Mortality counts were taken daily and the weight of the remaining rice grains were:recorded after 4 days. Triplicates were run for each test. The reduction in weight of rice compared to that of untreated rice indicated the antifeedant activity.

Antifeedant studies using E. hecebe
The 2nd instar larvae of E. hecabe (1.0-1.2 cm in length) were used in this study. The h a e were starved for 2 hours prior to the experiments. Fresh Kathurumurunga leaves containhg 10 leaflets (length range 3.8-4 em) were coated evenly with acetone/natural secretiodsyWic mixture diluted with acetone. Each leaf was placed in a small vial containing small mount of water to maintain turgidity. This leaf was put in a plastic bottle (2L) and a caterpillar was placed on the treated leaf. The plastic bottle was covered with a perforated cap for ventilation. Ten caterpillars were tested for each treatment. The caterpillars in the bottles with untreated leaves served as controls. After 112 and 24 hour periods, the leaves were examined and the number of leaflets eaten by the caterpillars counted. For the partially eaten leaflets, the mount eaten was recorded by the use of squared paper. The reduction in the amount of treated leaves compared to that of the untreated leaves indicated the antifeedant activity.

2.8.4a Preliminary tests for detertent activity
These tests were wried out following the method described by Eisner et a1. . 9 The test consisted of observing the response of a variety of insects to the vapours emanating from the tip of a capillary tube filled with the defensive secretion, presented to the insects manually at near contact range. A glass capillary tube filled with distilled water was presented to the insect first from m e side of the body (5 sec.) and then the other side (5 sec.). If the response to the control was negative (no movement towards or away from the capillary tube), a capillary tube f i e d with the natural secretion or synthetic mixture was presented (5 sec. on each side of body). The response was scored as positive if the insect turned away from the chemical under study both times.

Topical test for immtant activity
In nymphs of the cockroach P. americana, application (or near contact presentation) of a droplet of irritant to one side of abdominal tergite 4 or 5 evoked an accurately directed scratch reflex from the ipsilateral leg. The behaviour provides a basis for assay of topical irritancy of chemicals or their vapours.12. This technique was used in the present study to test the irritancy of the natural secretion and the synthetic mixture. Late-instar decapitated nymphs were used. Stimulation was affected by contact presentation (1% solutions in paraffin oil with 0.2 p1 droplets applied by microsyringe) or near contact presentation (glass capillary tube with pure sample held at less than 1 mm from the tergite surface). Imtancy of sample was scored as percent individuals that responded within 20 seconds of stimulation.

2.4.4~ Orfwtometer bioassay for repellent activity:
Two species of test insects 0. smaragdina and D. cingulatus were used in this study. A simple Y shaped insect olfactometer described by Gunawardena et al.1° was used with slight modifications (Figure 1). A capillary tube with the chemical under investigation (natural secretion or synthetic mixture) was kept in arm A. A capillary tube fded with distilled water was kept in arm B. A group of 20 test insects confined in the tube C was introduced to the main tube from the opposite end. These insects were made to move slowly towards the two arms of the olfactometer by gently warming the tube C (3S°C -4S°C). Proceeding towards the two arms, (A and B), insects encountered the air currents carrying the test compounds and made their selection of the arms (A or B). The number of insects in a particular arm was noted after 5 min. Batches of 20 insects were used at a time and each chemical was tested with four batches of insects. The test insects were not used repeatedly and parts of the apparatus likely to be contaminated were cleaned by washing and blowing air for 15 min before testing with a new batch of insects. The temperature of the laboratory during the period of this experiment was 28 + I0C and relative humidity 80 9 4%.

3.1
The Defensive Gland Structure ~6 e defensive gland of C. jams possesses a large bilobed, heartshaped median reservoir. The reservoir is reddish orange to dark red in colour. The gland is situated ventrally between the metathoracic legs and the reservoir extends back into the 2nd abdominal segment when distended with secretion. The two openings from the reservoir emerge on the ventral side near the anterior edge of the metasternum lateral to the mxal cavity. The wall of the reservoir has many fme cross striations and irregular folding which permit great distension. The secretion is yellow in colour. The volume of the secretion stored in the gland varies with the sex, the size of the individual and whether or not it has &charged recently. Not all the secretiosi is necessarily discharged on any one occasion. From a single individual 6-8 mg of fluid have been collected on some occasions.

2
Properties of Defensive Secretion and the Synthetic Mixture

As a topical applicant
24 hour LDso values for the test insects obtained by the method of topical application are listed in Table 2. Rice weevils (5. o-) and house flies (M. domeshca) were the most susceptible test insects (LWQ < lpg) to the defensive secretion. Addt cockroaches were not affected even at dosages upto 1 mg. Other test insects were moderately susceptible (lpg < ED% a 0.125mg). The synthetic mixture was found to be less potent than the natural secretion -of C. janus. All the test insects except cockroaches were moderately susceptible to the synthetic mixtuie. However the LBso values were somewhat higher than those of tht natuid secretion. 24 hour LC50 values for the storage pests obtained by fumigation method are listed in Table 3. Defensive secretion of C. janus and the synthetic mixture showed only a little promise as a fumigant against S. eve. b~a t u r a l secretion = defensive secretion of Cort&us@us, Synthetic mixture = t-2-hewenal : n-tridecane (60:40 w h ) , Number in parantheses indicate the 95% confidence limits.

As an antifeedant
The natural secretion of C. janus and the synthetic mixture were shown to be potent antifeedants to S. oyzae (Table 4) and E. hecabe ( Table 5).    b~esults are presented as mean 2 SD, n = 10, numbers in parantheses indicate the range of leaflets consumed by caterpillars, C d~e a n s ia one column, not followed by the same letter are significantly different p < 0.05 (ANOVA,Duncan' s Multiple Range test).
The antifeedant activity of the secretion of C. janus seems to be dependent on its dosage. The secretion and the synthetic mixture were tested with S. oryzae at three dosage levels. At the highest dosage (4 pg), only 7% of the treated rice was eaten. (The weight of untreated rice eaten by the weevils was considered as 100%). At a dosage of 0.04 pg, 30% of the treated rice was consumed. At the intermediate dosage (0.4 pg) activity was intermediate showing only 17% consumption. The mortality of the insects during the experiment was negligible (0.5%). However, 10-55% of the insects which consumed rice treated with natural secretion of C. janus or the synthetic mixture were dead after a period of 1M) hours.
At the dosages of synthetic mixture tested, 12-19% of the treated rice was eaten. The antifeedant activity does not seem to depend on dosage; no significant difference in the weights of rice consumed were found among the three dosage levels (P > 0.0.5).
Similar results (Table 5) were obtained for the antifeedant tests with E. hecabe. The defensive secretion and synthetic mixture were potent antifeedants to this species. At the highest dosage of natural secretion of C. janus (10%) only 4.5% of the treated leaflets was eaten while at 1% only 17% was eaten. At the dosage of 1% and 10% of the synthetic mixture, 19-20% of the treated leaflets was eaten.
Results of the p r e W a r y deterrency test proved that the natural defensive secretion and the synthetic mixture were repellents to i d the species tested namely  P. ameiicana, M. domestics, T. castaneum, S. oyim, 0. smaragdina and D. ci~rgukuius.
The topical irritancy test ( Table 6 ) showed that natural secretion and the syubetic mixture were equally active irritants to cockroaches (P >0.05). However, cantact presentation was more active than the presentation of vapour phase of the test materials. The results on the olfactometer bioassay for 0. smaragdina and D. cingrclatus ( Table 7) also showed that both the natural secretion and synthetic mixture were equipotent as repellents to both species tested.

Discussion and Conclusions
The present study has shown that the response of test insects, belonging to six different orders, to the two test materials under considereation, viz. the defensive secretion of C. janus and the synthetic chemicals mixture, depended greatly on the mode of application and the nature of the test insect. As a topical applicant, the defensive secretion was effective (LDso < lpg) against S. o y a e and M. domestic4 whilst other test insects showed poor sensitivity with the hardy P. mericana adults b e i i unaffected. Considering the chemical and physical differences in the cuticular frame work of test insects belonging to the six different orders, above variations are understandable.
In contrast, antifeedant property was very evident with both test materials under consideration. They showed almost equal effectiveness towards the two insect species tested, viz. S. oryzae (adult) and E. hecabe (caterpillar). The response of these insects to the test materials seemed to be dose dependent. Both insects showed high sensitivity to the test materials at low concentrations and more pronounced effects at higher concentrations.
The high antifeedant property of the defensive secretion of C. janus could be attributed to the process of 'self coating', common among insects which produce defensive secretions. Secretory discharges are usually effected by many Hemipterans in a diversity of ways. 798, 16 Usually the defensive secretions characteristic of a particular insect species is an optimised chemical formulation suited not only for chemical deterrence of the enemy but self-coating which prevent it being eaten or picked up by predators. Insects which are vulnerable to such predators are known to use this method as a means of self protection. 12 Fumigant properties of the test materials were tested against two species (storage pests) only vlz. S. oryuIe and T. cartuneurn. Results do not appear to be remarkable (LC% c 10 for S. aye and LC% c 20 for T. cartmeurn), but this could be a valuable result in conjunction with the high antifeedant properties of the above two materials towards the storage pest, S. o y m .
Considering the deterrent properties, the defensive secretion as well as the chemical mixture showed equally high effectiveness. Repellent properties also do not differ greatly between the two although the natural defensive secretion was found to be superior. Both test materiah were found to be potent irritants to the nymph P. ameiicana. It is known that secretions of terrestrial Heteroptera are primarily defensive irritants14115 and hence high irritancy of the defensive secretion is not surprising.
The data showed that the defensive secretion of C. janus and its corresponding chemical combination afforded protection against insects. This result is not s u r p r i s i when the biological aspects of this insect are considered. C. janus is a large, crawling insect which hardly ilies. Its vulnerability to predators such as birds and ants found in the sweet pumpkin fields is high, and production of a toxic defensive secretion by C. janus is therefore, not surprising.
A marked parallelism in activities of the defensive secretion of C. jmus and its two major volatile constituents t-2-hexenal and n-tridecane, (60:40 wtw) were shown in this study. This is an important finding for practical application of the findings for control of insect pests. One of the major problems that could arise is the difference in volatilities of the two components of the mixture. Under field conditions there could be a greater loss of more volatile t-2-hexenal, resulting in an altered composition. It has been shown previously'1, that the activity of the above mixture is very dependent on the ratios of the two components, with the 6&40 (w/w) mixture of t-2-hexen* n-tridecane being the most effective. Consequently the use of field sprays for the control of E. hecabe may pose problems, although in the w e of storage pest, S. o w , application of the above mixture as an antifeedant will be in a confined area, and hence the chances of it being successful are much higher.
The following conclusions could be drawn from this study. The effectiveness of toxicants, viz. the defensive secretion of C. junus and the chemical combination, t-2-hexenal and n-tridecane (60:40 w/w) depends largely on the mode of application and the type of test insect. A parallelism was seen in the activities of defensive secretion of C. junus and the combination of its two major volatiies t-2-hexenal and n-tridecane. Among the two test materials however, the natural defensive secretion is slightly more effective than the synthetic mixture as a topical applicant, a fumigant and an antifeedant while both are equally highly active as irritants and repellents.
The most effective modes of activity of test materials were as antifeedants, irritants, and repellents. S. o w e and E, hecabe had high sensitivity to antifeedant activity, whilst nymphs of P. mericuna were markedly susceptible to irritants and 0. srnmagdina and D. cingulatus were fairly sensitive when used as repellents.
In view of the practical implications, the chemical combination t-2-hexenal and n-tridecane could be a satisfactory substitute for the defensive secretion of C. janus. Finally, taking into account the volatile nature of t-2-hexenal in the above mixture together with the data obtained, two practical applications could be recommended as plausible, viz. the use as an antifeedant against S. oryzae and as an irritant against P. americana.