HOST ATTRACTANTS FOR THE BANANA STEM BORER, Odoiporus longicollis (COLEOPTERA: CURCULIONIDAE): IDENTIFICATION, ELECTROPHYSIOLOGICAL ACTIVITY AND BEHAVIOURAL BIOASSAY

Steam volatiles ol'freqj~ly cut hanana stem were analyzed lly comhined . gas cbromato~~apl~yelectroai~l;enno~apl~ic detection (GC-EAD) ailtl comhined gas cl~i~c~matograplly-niass spect~.ometry (GC-MS) to iclentifi n-hexan.ol, n-hcxanal, n-pentanol and cis-3-hexc?nol as host attractants for the banana stem borer, Cldotl~oi~c~s loir,qicollis. .Idcnlities of t h e ahovc compounds were confirmed 1,y comparalivc studies with aulhcntics. Female weevils responded sligl~tly mo1.c (:EAG 1.41k !).()I) t1:han male weevils (EAG 1.04 f 0.03) to tho steam distillate. Synthel;ic eclu~.valents of the ahove aktractant,~, were sul?iect;ed to EAG assay and 11-hexan0l elicitcd the l~iglicst EAG lesponse (0.39 k 0.02 mV)'in Ihc antelma c l t ' tllis weevil. The other synthetic at.Lractants, 11-hexanal, n-pentanol and cis-3-hexello1 elicited EAG responses in t.ba fi~llowing order; 0.15 mV, 0.10 mV, 0.0t) niV. Aniix(;u~.c ofthe ahovc synL11etic attraclants (SAM) also elicited a poor EAG response (0.2 nlV) in. O.loii,gicolli,<. In lield ,xpcljments n-hexanol singly o]..,in comhiuation with the agglogation phcromone o.TO, lorrgicollls Cailcd tc.) attract any weevils into traps. The synthetic nttractanl; mnix(;lli~e (SAM) also was not field nt.tractive while it,s combination wit11 lhe aggregation pl1er1)nlone of 0. loitgicollis was attractive wit,h a maximum rrleail catch of' ().(Xi k 0.21 weevils ILrapIweek. Undei the same contlitions, Ilowcvc~; thepest's known atlraclant system, the aggregation phcromone + banana stem tissue haited trap, 11ad a maximum mean catch of 29.16 + 7.02 weevils1 traplweek. K e y words: Banana stem horel; n-liexanal, 11-hexanol, cis-3-hexenol, host at.tractaiits, Otloil~ori~s h>~i.g'fiil:Olli~ and n-pentarrol.


INTRODUCTION
Banana (Family: Musaceae) is one of the most important fruits j n the South East Asian region. A large number of banana clones used over the world derive originally from two wild species, Musa acum iuata and Musa balhisiana.' At present, Sri Lanka has 29 clones of banana of which many are indigenous to the country. The country has a favourable climate for banana and a total of about 45,000 hectares of banana has been recorded for the year 1995.2 Banana cultivators in the region are troubled by two weevils, the banana stem borer, Odoiporus longicollis' and the rhizome borer, Cosnropolite sordidus."." Female 0. longicollis lay eggs in the stem (pseudo stem) and bunch stallts of the banana tree. The larvae tunnel through the stem tissue, making j t weak and liable to get blown down. This pest is controlled by chemicals such as DDT, BHC, aldrin and dieldrin used as the soil treatments with repeats every 6-12 months. In some areas the base of the banana stem is painted upto about 50 cm with tar or mud slurries containing insecticides to prevent attack by 0. longicolli~.~ A liberal spraying of insecticides . in the insect's hiding sites has been found to be an efficient method t o control this pest." In some areas of Sri Lanka, mass trapping of this pest using split banana pseudostem is being practiced. The only biological control method reported for banana weevils is the use of predators and parasite^.^ Studies on attractants for banana pests have been limited mainly to aggregation pheromones. Beauhaire ct.al. identified the aggregation pheromone of C. sordidus in 1995 as a terpenoid ketal.Vhis was shown to be field active only in the presence of host tissue. We earlier reported the identification of male aggregation pheromone of 0.lorrgicollis as 2-methyl 4-heptanol."~" This pheromone was also found to be field active only jn combination with banana stem tissue." The present paper reports the identification of several host attractants from banana stem tissue. Attempts have been also made to use synthetic equivalents of the identified attractants as field baits. In this study we have used a banana cultivar widely grown in Sri Lanka (Seeni Kesel) a triploid (ABB) whose volatiles were highly attractive to the above pest.

I I L S C C~S .
Adult 0. 1on.gicollis were collected from rotten banana plants in the District of Gampaha, Sri Lanka. They were separated by sex, kept at a t,emperature of' 25 f 2 "C and Rh of 80 f 4 (5, and fed with banana stem tissue. Steam Distillatiol/: Fresh banana stem cut into small pieces (1Kg) was steam distilled for 4 h in an all glass apparatus. The distillate (110 ml) was saturated with NaCl and subsequently extracted wit11 diethyl ether (30 ml x 3). The ether phase was dried over anhydrous magnesium sulfate and concentrated down to 2 ml. This solution was kept at 4 "C to be used later in bioassay and analytical studies. Using the above procedure, steam distillates of four varieties of banana, all triploids viz Seeni Kesel (ABB), Ambul Kesel (AAB), Ananlalu (AAA) and Kolikuttu (AAB) were prepared for a bioactivity comparison.
Behavior~ral Bioassay: The attractant properties of banana varieties were compared by a behavioural bioassay using a Y-shaped glass olfactometer.'"he olfactometer was covered with black paper in order to facilitate free movement of weevils inside the tube. Appropriate doses of steam distillates were absorbed into filter paper strips and solvent was allowed t,o evaporate. This was next inserted in to one arm (e.g. A) of the olfactometer. Blanks were prepared from filter paper strips containing similar amounts of diethyl ether and inserted into the other arm (e.g. B) and this sequence was interchanged in subsequent replicates. Weevils, starved 3 h before the bioassay, were released into the third arm of the 01.factometer gradua1,ly. Weevils that did not move into the baited arm within 3 mi.11 were consj.dered non-responders. Each test substance was assayed using 6 batches of weevils each consisting of 5 males and 5 females.
An,alysis of volatiles: Steam volatiles were analyzed on a Hewlett Packard 5890 series I1 chromatograph equipped wit11 fused silica capillary columns (30111 x 0.25mm-ID 0.25 pm fjlm thickness, J & W Scientific California) coated with DB 5 stationary phase, a flame ionization detector (FID). Helium was the carrier gas a t a flow rate of 1 mllmin and nitrogen at 30 mllmin, the make up gas with a temperature program: 40 "C (5 mjn) from 40 -170°C (rate 5'C/min), from 170°C to 250°C (rate 20°C/ min) and finally at 250% for 10 min.
Gas chromatography-mass spectrometric (GC-MS) analyses were performed with a Hewlett Packard 5890 Series I1 gas chromatograph, fitted to a HP 5870 Series mass spectro~net~er with electron impact (EI) mode at 110 eV. GC column parameters were similar to those described previously, however, with tlze following temperature program; 40% (5 min), from 40-170°C (rate 5 T / min), from 170 -250°C (rate 20'CImin) and a t 250 "C for 10 min.
Electl-otr ~zteruzograplzy : Electroaritenriograms (EAGP1 were recorded from isolated antennae of male or female 0. longicollis using glass capillary electrodes filled wit11 saline solution and connected to gold wire. The reference electrode was inserted into the base of the antenna and the recording electrode into the antenna1 cone. The electrical signal was fed into an amplifier and displayed on a computer screen (Gateway 2000).
These compounds in the above proportions were found to occur in the banana steam volatiles For the above experiment both male and female weevils were used. Each sailiple was tested on six antennae (male and female). An external source of' 1 inV was used to calibra1,e the EAG response. As correction for ;~nt,ennal fatigue, t l~e response to reference substance (100 pl of steam distilled extract of banana) was measured between test samples. Each dose was delivered onto a filter paper s t r~ p ( 3 X 30 mm) placed Inside a glass Pasteur pipett,e (0.5rn11.1 ID X 100mm).
To record the dose response curves ofnlale and female 0 , longicollis, eight doses (1, 5, 10, 50, 100,200,300, 500 p1) of the banalla steam distillate were t,ested. Each dose was delivered as 100 pl aliquots in order to maintain a constant volume in samples. Thus doses having less than 100 pl were diluted whilst those which had a volu~rle more than that were concentrated until the sample size reached 100p1.
Covpled gas ch,romatograph,y-eZectroa17~te17~nographic detection.: GC-EAD was performed with the same GC described earlier but with a splitter fitted into the column. Effluents were split between FID and antenna in the ratio 20: 80 respectively. A make up gas of nitrogen (30 mllmin) was added to the splitting device to accelerate the exit of GC effluent towards the antenna. The effluents from chromatograph were driven through a heated transfer line (200 'C) in order to prevent any condensation. The EAG and FID signals were synchronized using a GC-EAD response of R. ferrr~gin,eus antenna to its aggregation pheromone.
Fielcl assay: Following the method described previo~~sly,~%eld experiments were conduct,ed in banana plantations in Kadawata in the District of Gampaha from Mag-July 1997. White plastic buckets (capacity 5L, diameter 21cm, height 17 cm) filled with soapy water (1.5 L) were used as traps. Buclret traps were attached at 3-4 feet height to strong wooden poles (posted near the banana plant) in randomized bloclrs with traps a t 20m intervals and blocks 1 km apart. Six candidate baits and two controls (see later) with 6 replicates each were tested. Attractants filled into glass capillaries (20mg i n each) were suspended inside the bucket just above the water level. Rate of evaporation of attractants from similar capillaries was determined to be 400 ~~gfday. One end of the capillaries was opened at the site, prior to the start of the experiment. Test baits used in traps were, n-hexanol, the synthetjc attractant mixture (SAM) and SAM in combjnation with the synthetic aggregation pheromone, 2-methyl-4-heptanol!' (1:1, wfw). In preparing the above combinations, attractants were not mixed into one capillary but were kept jn different capillaries. As the two controls, a blank ( a trap with no attractants) and a lrnowrl attractant systern" were used. T11e latter consisted of the aggregation pheromone + srnall pieces of fresh banana stem tissue (800g) placed inside t,he bucket, (in water). Trap cat,clles were counted weekly for 6 weeks.

Bioactivity of steam distillates
EAG act,jvity EAG response profiles of both female and male 0. lollgicollis to the steam distillate were similar, female responses being slightly higher than those of the male. The highest EAG responses were observed in male and female antennae at EAGs of 1.04k 0.03 and 1.41 + 0.01 mV respectively at a dose of 100 p.1 (Figure 1).

Rehavioural assay of the host volatiles of banana varieties
Volatiles of Seeili Kesel elicited the highest activity by attracting a mean number of 7 weevils out of 10 into the baited arm. This, however, was not significantly different from the mean numbers of weevils attracted by Anamalu and Kolikutt,~. The weakest attractant was the volatiles ofAmbul Kesel which had a mean of 3.66 weevils in the baited arm ( Figure 2). Bioassay using seven doses of the Seeni Kesel steam distillate showed a maxjmum a t 10 111 (Figure 3).   the cai-riar gas at a flow i.atc+ of 1 mumin and nit:rogen a t 30 mVlnin, the make up gas with a 1emperai;ure proRrnrn: 40 "C (5 mini born 40 "C -170 "C (ra(.e 5 "Clmin), kom 170 'C to 260 "C (v:lb 20 "C/ min) and finally at 250 "C for 10 min. GC-MS analysis: Acomparative study of the mass spectra of the GC peaks A,B,C,D and those of the respective standard compounds identified them as pentanol, hexanal, cis-3-hexenol, and hexanol respectively. Subsequently GC retention times of the above pealrs were compared with those of the standards (Table 1).
Comparative EAG assay of synthetic host attractants (individual), a mixture (SAM) and the steam distillate (natural host): The highest EAG response ( Field bioassay: Traps containing hexanol alone, hexanol + aggregation pheromone or SAM alone did not attract weevils into traps. Traps baited with SAM + aggregation pheromone caught weevils with a variable weekly average, the maximum being 0.66 f 0.21 weevilsltrap. The control with t h e known attractant system, aggregation pheromone +banana stem tissue continued attracting weevils for 6 weeks with the highest weekly average catch of 29.16 f 7.02 weevilsltrap. ( Table 2). The average weekly trap catches for the above two traps over 6 weeks ( 0.25 SD? 0.24 and 11.58 SD+ 10.27 respectively) were significantly different (P<0.05, one way ANOVA & Scheffer's test). In all traps approximately equal number of male and female weevils were caught (male: female=43:57).

DISCUSSION
Host volatile synergism with aggregation pheromones has been established for Curculionids in recent years.'", l4 The host tissue + aggregation pheromone combination has always been a better lure for a particular pest species compared to that of aggregation pheromones + synthetic host volatiles. However, difficulties in the nlaintenance ofhost tissue in pheromone baited traps for long periods has made the above practice less popular. One good example is the aggregation pheromone baited taap of' the date and coconut weevil, Rhy~1chophoru.s ferrugineus. Host volatiles of toddy, pieces of apple, coconut petioles, sugar cane etc. have been used successfully in the above pheromone trap, but with the need of weekly replacement. In extensive weevil contxol programs in the Middle Eastern date palm fields where several thousands of t-raps are deployed, the above traps have not shown much promise (unpublished observations of N.E. Gunawardena). In these instances alternatives have been the synthetic attractants having synergism with aggregation pheromone. Recent work on synthetic synergists for Curculionid aggregation pheromones has led to a large number of attractants. For example kiromonal synergism of ethyl alcohol for R. ferrr~gin,eus,'%thylbutyrate for West Indian sugar cane weevil (WISW), ethylacetate for Rhynchophorus cru.entatus and R. palmarum, ethylpropionate for R. phoendcis have been r e p~r t e d .~W s u a l l y these compounds have been components of the respective natural host volatiles but their synergism has not been proximate to those of t,he respective host plant activity. In some cases, the host volatile synergism h a s been enhanced by the addition of synthetic attractants.
Our previous study showed that the synthetic aggregation pheromone baited t-rap was effective for 0. 1on.gicolli.s only if banana stem tissue was present."." Replacement of tile banana stem tissue with a synthetic chemlcal or a mixture of chemicals was thought to open wider field applications for the above trap. In our preliminary laboratory behavioural bioassay, banana stem tissue alone was found t,o be attractive to the weevil, albeit a t a short range (70% weevils attracted to baited arm). In an EAG assay the banana steam distillate created similar response profiles for males and females but with a stronger response from the latter. This probably is indicative of the host volatile's plausible additional function as an oviposition stimulant for females.
Our identification of n-pentanol, n-hexanol, n-hexanal and cis-3-hexenol as host attractants for 0. lor1,gicollis is in conformity with the previous findings for Curculionids. Further, the Coleopteran's attraction to short chain aldeh$des, acetates and alcohols has been well doc~mented.1"~~ EAG results of the present study indicated only hexanol as the potent synthetic attractant for the banana weevil (EAG 0.39 mV). However our field work proved that hexanol is neither a long-range attractant nor a synergist to the aggregation pheromone. In contrast, the synthetic attractant mixture (SAM) proved t o be a synergist for the aggregation pheromone althougll poorly. Under the same field conditions it was also shown that the host tissue is a much better synergist for the aggregation pheromone. This suggests two possibilities: the presence of additional, as yet unlrnown banana stem kairomones or the production of different chemicals in the rotting banana stem tissue. In either case further work is necessary.
In conclusion, firstly, if there are undetected host kiromones present in the host tissue, it could be overcome by using a different method to collect banana stem volatiles. The presently used steam distillation method is often not the best wag to collect volatiles for GC-EAD analysis although it was the easiest method available to us. Porapalc or headspace volatile collections are superior to steam distillation although the requirement of substantial amounts of volatiles by the former is a challenge. Secondly, rotting banana sten1 could be examined for new attractants.