Rothamsted Repository Download

23 BACKGROUND: The banana weevil, Cosmopolites sordidus , has been frequently cited as the 24 most challenging constraint to banana and plantain production, particularly in small-scale 25 (smallholder) farming. For the development of a new, low-cost weevil management technology 26 based on attractive host plant material, we previously identified (2 R ,5 S )-theaspirane as the 27 active component of attractive senesced banana leaves. In this new study, we used behavioural 28 (olfactometer) bioassays with adult weevils to compare the attractiveness of four different 29 developmental stages of banana leaves, ie. unfolding (pale green), matured green (deep green), 30 matured yellowing and senesced, to determine which leaf developmental stage would be most 31 appropriate for use in weevil management. We also investigated the attractiveness of senesced 32 leaf extracts prepared using different solvents to determine which solvent would be most 33 appropriate for local production of leaf extracts. chromatography-34 electroantennography (GC-EAG) was then used with adult weevils to confirm the presence of 35 (2 R ,5 S )-theaspirane in attractive leaf extracts. RESULTS: Of the leaf materials tested, only the odour of senesced leaf material was 37 significantly attractive to adult weevils ( P <0.005). Furthermore, an extract of senesced material 38 prepared using palm wine alcohol was significantly attractive ( P <0.05). Using coupled GC-39 EAG with weevil antennae, (2 R ,5 S )-theaspirane was identified as a minor component with 40 strong EAG activity within the palm wine alcohol extract. CONCLUSION: The results suggest that palm wine alcohol extracts of senesced banana leaf material could be used to lure adult C. sordidus to traps in the field, as part of an ethnobotanical-43 based approach for C. sordidus management on smallholder farms.

distilled water to provide a moist environment for the weevils. The weevil culture was kept at 114 room temperature and inspected daily to remove dead weevils, whilst the tissue mats were 115 replaced or moistened weekly. The weevils and containers were also cleaned using running tap 116 water, and the feed rhizomes replaced every month 36 . Weevils used in the experiments were 117 adult unsexed members taken from within the weevil culture, placed in a container without 118 food (rhizome) and starved for at least 12 hours before use.

Volatile collection.
Volatile organic compounds (VOCs) from senesced leaf material were 121 collected by air entrainment (dynamic headspace collection) using standard techniques 37 . Air 122 filtered through activated charcoal was pumped into a glass jar (5 L) containing 80-100 g of 123 senesced leaf material at a rate of 900 ml min -1 and the VOCs trapped on Porapak Q (50mg,   chamber was left empty to serve as a blank control. In choice tests, the same mass of leaf 159 material was placed in each of the test chambers. In tests using VOC samples, 20 µl of VOC 160 extract (equivalent to 0.08 g leaf material) was applied on equally sized 1cm 2 pieces of 161 Whatman® qualitative filter paper, Grade 1 (Sigma-Aldrich, UK) in each test, and the solvent 162 allowed to evaporate (ca. 10 sec) before use. Filter papers with solvent were used as controls.

163
In tests using solvent extracts, 20µl equivalent quantities of extract and pure solvent were 164 applied on the filter paper. The solvents were allowed to evaporate (ca. 10 sec) before the pieces 165 of paper were placed in the olfactometer. For each assay, groups of ten unsexed, adult weevils 166 removed from the weevil culture were used 36 . Test weevils were placed in the main chamber, 167 and allowed 20 -30 minutes to respond to and move towards stimuli in either response chamber. After the first test period, if two (20%) or more weevils failed to respond to either 169 test material, a further 10-15 min was allowed for them to respond. At the end of the permitted 170 time, the numbers of weevils in each chamber were recorded. For each experiment, a total of 171 200 different weevils were used in twenty (20) replicates. The test materials were switched 172 between chambers midway through the experiment (after first ten replicates). Before the switch 173 of test materials, the apparatus was cleaned with dilute ethanol, thoroughly rinsed in distilled 174 water, wiped dry with tissue paper and allowed to air dry for 30 mins.    to the antennal preparation and the GC detector, has been described previously 41 . Separation 191 of the collected palm wine alcohol extract was achieved by high resolution gas chromatography

210
In olfactometer assays, both freshly collected and freshly collected, laboratory-dried samples 211 of banana leaf material, from different developmental stages, ie. unfolding (pale green) banana 212 leaf, matured green (deep green) banana leaf and matured yellowing banana leaf, were not 213 statistically more attractive to adult C. sordidus than the clean air control (Table 1). Senesced 214 banana leaf material, however, was significantly more attractive to weevils than the clean air than the dried unfolding banana leaf sample (P<0.03, Table 1). An extract of volatile organic 217 compounds (VOCs) collected from senesced leaf material by air entrainment was significantly 218 more attractive to adult weevils than a solvent control (P<0.001, Table 1). In choice assays, 219 there was no significant difference between the attractiveness of senesced leaf material and 220 collected VOCs (Table 1) to banana weevils. A solvent extract of senesced leaves prepared 221 using palm wine alcohol was significantly more attractive when compared to a solvent control, 222 whereas extracts prepared using methanol, ethanol and hexane did not differ from solvent 223 control in attractiveness to the test weevils (Table 2). There were no statistical differences in 224 attractiveness of extracts made at ambient temperatures and those extracted using Soxhlet 225 apparatus (Table 2). When compared to whole leaf material, only the methanolic extract was 226 significantly attractive (Table 2). In olfactometry assays to ascertain the comparative 227 attractiveness of extracts of senesced leaves in the various solvents to C. sordidus, palm wine 228 alcohol extract was similar to those made in methanol and hexane but was significantly more 229 attractive than the ethanol extract (Table 2). Coupled GC-electroantennography (GC-EAG) 230 analysis, on a non-polar DB-1 GC column, of the palm wine alcohol extract (the more attractive  olfactometer bioassays, whereas fresh plant material, including wounded tissue, was not attractive 9-11 . Furthermore, senesced leaf material was reported to be more attractive when 241 compared with banana rhizome and pseudostem, cocoyam and dead grasses 9,10 . In this study, 242 banana leaves that were harvested as fresh materials and those artificially dried in the 243 laboratory, ie. unfolding (pale green) leaf, matured green (deep green) leaf and matured 244 yellowing leaf material, were not significantly attractive, whereas senesced leaf material was 245 significantly more attractive compared to clean air (Table 1). Comparing the three types of 246 artificially dried leaf material to senesced leaf material, significant attraction to the senesced 247 leaf material was only observed when compared to the dried unfolding pale green leaf material.

248
There was no significant difference in attractiveness between the senesced leaves, the dried