J Chem Ecol DOI 10.1007/s10886-014-0512-3
Plant Volatiles Influence the African Weaver Ant-Cashew Tree Mutualism Caroline Wanjiku & Fathiya M. Khamis & Peter E. A. Teal & Baldwyn Torto
Received: 21 May 2014 / Revised: 22 August 2014 / Accepted: 27 August 2014 # Springer Science+Business Media New York 2014
Abstract Plant volatiles influence virtually all forms of antplant symbioses. However, little is known about their role in the mutualistic relationship between the African weaver ant and the cashew tree. In this study, we tested the hypothesis that cashew tree volatiles from plant parts most vulnerable to herbivory viz. inflorescence, leaves, and fruits, are attractive to weaver ants. Using behavioral assays, we show that these volatiles attract weaver ants but without significant difference in preference for any of the odors. These same plant parts are associated with extra floral nectaries (EFNs’) and therefore we evaluated the possibility that the ants associate the volatiles with food rewards. We found that perception of the odors was followed by a searching response that led the ants to nonvolatile sugar rewards. More importantly, we observed that weaver ants spent significantly more time around the odor when it was paired to a reward. Chemical analysis of volatiles showed that the plant parts shared similarities in chemical composition, dominated by monoterpenes and sesquiterpenes. Additionally, we evaluated the attractiveness of a synthetic blend of three ocimene isomers ((E)-β-ocimene, (Z)-βocimene and allo-ocimene) identified in cashew leaf odor and shown to constitute a candidate kairomone for the cashew pest Pseudotheraptus wayi. We found that the attractiveness of the blend was dose dependent, and the response of the ants was not significantly different to that established with the C. Wanjiku : F. M. Khamis : B. Torto (*) International Centre of Insect Physiology and Ecology (icipe), P.O. Box 30772-00100, Nairobi, Kenya e-mail: [email protected] C. Wanjiku : F. M. Khamis Department of Biochemistry and Biotechnology, Kenyatta University, P.O. Box 43844, Nairobi, Kenya P. E. A. Teal USDA/ARS-Center for Medical, Agricultural and Veterinary Entomology, Gainesville, FL 32608, USA
crude volatiles from plant tissues. These results present new and interesting possibilities for improving weaver ant performance in cashew pest management. Keywords Oecophylla longinoda . Pseudotheraptus wayi . Mutualism . Olfactory cues . Plant volatiles . Cashew . Pest management . Hymenoptera . Formicidae
Introduction Ants are known natural enemies of many different insect pests, and their use dates back hundreds of years (Heil 2008). In Africa, Oecophylla longinoda (Latreille) (Hymenoptera: Formicidae), also known as the African weaver ant stands out in this realm. Having been discovered as early as 1953, this ant was and continues to be especially useful in the biocontrol of insect pests of cashew (Anacardium occidentale), coconut (Cocos nucifera), mango (Mangifera indica), and cocoa (Theobroma cacao) all across Africa (Ayenor et al. 2007; Dwomoh et al. 2008; Olotu et al. 2013; Van Mele 2008; Van Mele and Vayssières 2007; Van Mele et al. 2007; Way and Khoo 1992). Ant-plant interactions take many different forms but can be c a t e g o r i z e d b r o a d l y in t o tw o g r o u p s ; d i s p e r s a l (myrmecochory) and defense mutualisms that are either facultative or obligate (Heil 2010; Youngsteadt et al. 2009). Facultative mutualisms, are built on the exchange of food rewards; elaiosomes (specialized nutrient rich seed structures) in the case of dispersals, and food bodies or extrafloral nectaries (EFNs) in defense associations (Heil 2010; Heil and McKey 2003; Koptur and Truong 1998; Turner and Frederickson 2013). Obligate mutualisms on the other hand are based on provision of nesting sites for both dispersal and defense symbioses (Boucher et al. 1982; Heil 2008, 2010; Youngsteadt et al. 2009).
J Chem Ecol
The perennial association between the cashew tree and O. longinoda also is built on the provision of food rewards from EFNs in exchange for protection from herbivores (Rickson and Rickson 1998). The nectar secreting organs are located on panicle branches, leaves, and young fruits, which are also the most susceptible parts to herbivory (Way and Khoo 1992; Wijit 1991). The tree also accommodates honeydew producing hemipterans that are equally exploited by ants for honey dew (Way and Khoo 1992; Wijit 1991). As such, a potential defense mutualism is suggested (Rickson and Rickson 1998). Plant volatiles have been shown to mediate dispersal and defense mutualisms. In dispersals, responses of ants to seed odors (Youngsteadt et al. 2008, 2010) as well as volatiles emanating from elaiosomes have been reported (Gammans et al. 2006; Turner and Frederickson 2013). In defensive associations, ants also have been shown to respond to herbivore induced volatiles (Agrawal and Dubin-Thaler 1999; Bruna et al. 2004). At this point, the chemistry of herbivore induced damage in the cashew tree is still not clear. However, a recent study by Egonyu et al (2013), demonstrated that a blend of three components ((E)-β-ocimene, (Z)-β-ocimene, and allo-ocimene) in cashew leaf odor constitute a kairomone that elicits short-range attractive response from males of the cashew pest Pseudotheraptus wayi. This pest belongs to the order heteroptera; a group of sap sucking bugs (Aldrich 1988) ranked fourth among the most economically important groups of insects (Millar 2005). Besides cashew, P. wayi also attacks coconuts, avocados, macadamia, guava, and mango among other crops (Olotu et al. 2013). So far, the use of the predatory African weaver ant remains the most promising management strategy for the control of P. wayi (Brown 1955; Olotu et al. 2013; Vanderplank 1960), and this links small holder farmers to lucrative organic markets (Van Mele 2008). In view of this, it is important that the chemical ecology of the cashew tree system be investigated. We sought to understand how volatile organic compounds (VOCs) influence weaver ant behavior. We hypothesized that cashew volatiles emanating from the plant parts deemed most vulnerable to herbivory are attractive to the African weaver ant. To test this we carried out studies to determine: (a) the response of weaver ant workers to crude volatile extracts from cashew leaves, fruit, and inflorescence; (b) the likelihood that weaver ants relate the cashew odors to a food item, given the association of the test plant parts to EFNs; (c) the chemical composition of the volatiles and how the compounds and quantities compare to each other; and (d) the responses of the ants to a commercial blend of three isomers of ocimene, (E)-β-ocimene, (Z)-β-ocimene, and allo-ocimene that attract male Pseudotheraptus wayi. We used worker ants for this study because they play a significant role in defense and foraging (Heinze et al. 1994; Schwander et al. 2005) and in the biological control of cashew pests.
Methods and Materials Insects The weaver ant colony was established from 12 nests collected from the Mtwapa area (3.9500° S, 39.7444° E) in Kikamabala divison of Kilifi district in the coast province of Kenya in September 2012, containing approximately 200 or more ants per nest. The insects were transported to an outdoor insectary in the Animal Rearing and Containment Unit of icipe, Duduville, Nairobi campus. They were inoculated on mango seedlings (≈3½ft high) and allowed to habituate and weave fresh nests in the insectary at 24-28 °C and 55-65% relative humidity. They were maintained on a diet of 6% sugar water and powdered fish meal (silver cyprinid- “omena”) as a protein source purchased at a local store. Experiment 1: Olfactometric Assays with Cashew Volatiles A Y-tube olfactometer (arm length: 7×7 cm, internal diam: 1 cm) was used with one arm connected to the test odor; crude extract of cashew volatiles collected from leaves, fruits, and inflorescence, each dissolved in 25 μl dichloromethane (Analytical grade -Sigma Aldrich, 3050 Spruce street St. Louis, MO, USA) and loaded on a filter paper disc (3×3 cm Whatman filter paper No. 1). The other arm was connected to a control (a similar volume of solvent loaded onto a similar size and type of filter paper disc) via Teflon tubes. The crude extracts were tested in a dose response assay at 0.3, 0.6, and 1.2 plant hour equivalent (PHE) where 1 PHE=volatiles emitted by the part sampled on each individual plant per hour. An energy-saving bulb (Philips stick -11 W) emitting white light was suspended 55 cm above the bioassay arena to provide illumination. Charcoal-filtered clean air was passed through the Teflon tubes into each arm of the olfactometer at a flow rate of 348 ml/min and pulled out of the main arm of the olfactometer at the same rate by a battery powered portable vacuum pump (supplied by co-author Peter E.A. Teal; USDA/ARS-CMAVE, Gainesville, Fl, USA). Individual worker ants were released in the main arm of the Y-tube and allowed to settle for 5 min. The pump system was turned on to test the ant preferences for either the test or control odor, and each ant was given 5 min to make a choice. After every 5 insects, the odor laden filter paper and Y-tube were routinely replaced. However, when the test insect released a droplet of brown liquid from the anal region, indicating a trail laying response, the Y-tube was immediately replaced. The positions of the test and control arms of the Y-tube also were switched after every 5 insects to minimize positional bias. Preference was determined when the insect traversed the entire stem of the olfactometer and an additional 2 cm into either arm. The total time spent at each arm was recorded. Extracts of volatiles from each plant part were tested against 10 individual ants and replicated five times in each case.
J Chem Ecol
Experiment 2: Cashew Volatiles Paired with Sugar Rewards This bioassay was carried out in a glass petri-dish arena, 15 cm wide x 3.5 cm high lined with a circular wire mesh screen (mesh size≈1.19 mm) and divided into four equal quadrants. Two quadrants were designated as test and control, whereas the remaining two were left as neutral zones. The test comprised extracts of volatiles from leaves, fruit, and inflorescence (25 μl, 0.6 PHE) loaded on rubber septa. The control was rubber septum loaded with only 25 μl dichloromethane. Both septa were left to stand for about 10 min to allow the solvent to evaporate. The extract-loaded septum was suspended on the wire gauze in the test quadrant of the bioassay arena with a pin, and the other in the control quadrant. Ten droplets of glucose syrup (2:1 glucose in water) were applied on the wire mesh in both test and control quadrants using a fine camel hair brush. At this concentration, each droplet of the sugar syrup was held on the mesh screen without flow. An ant held in a small glass vial (21×34 mm) then was placed at the center of the bioassay arena and allowed to crawl out of the vial into the arena, after which the vial was removed and the arena covered with its glass cover. Each ant was given 10 min to walk around the arena, and the time spent as well as the frequency of sugar feeding in the test and control quadrant was recorded. After every 5 ants, the rubber septa were replaced and their position in the petri dish switched to minimize on positional bias. In the event of a trail laying response, the petri dish was replaced immediately before testing subsequent insects. Each odor, paired with sugar rewards, was tested against 10 individual ants and was replicated five times in each case. Experiment 3: Collection and Analysis of Cashew Volatiles Volatiles released from cashew were collected from plants growing at the Kenya Agricultural Research Institute (KARI) orchards in the Mtwapa area of Kilifi district in the coastal region of Kenya (3.9500°S and 39.7444°E). Volatiles were collected from leaves, fruit, and inflorescence on Super Q (30 mg, Analytical Research Systems, Gainesville, FL, USA) and Supelco SPME fibers (65 μm polydimethysiloxane-divinlybenzene, Sigma-Aldrich, Supelco Park, Bellefonte, PA, USA) for comparative purposes. About 12–15 leaves, 9–12 fruits, and 10–15 inflorescences were bagged in oven-baked plastic bags (355 × 508 mm, Classic Consumer products, Inc, Englewood, NJ, USA) separately, and volatiles were collected from these sources for 12 h (Super Q) and 8 h (SPME) . Each plant part was sampled four times using a different plant in each replicate. Volatiles trapped on Super Q filters were eluted using 500 μl dichloromethane (Analytical grade, Sigma Aldrich, St. Louis, MO, USA) and stored at −80 °C until use for chemical analysis and bioassays. Each individual sample contained 12 P H E . Vo l a t i l e s w e r e a n a l y z e d b y c o u p l e d g a s chromatography/mass spectrometry (GC/MS) on an Agilent
technologies Series A 7890 GC coupled to a 5975C (inert XL/ EI/CI MSD) triple axis mass detector, equipped with an HP5MS column 30 m×250 μm×0.25 μm in the electron impact mode at 70 eV. The GC oven temperature was 35 °C for 5 min with a rise of 10 °C/min to 280 ° C for 10.5 min, then 5 °C/min to 285 °C and held at this temperature for 9 min. Identification of compounds was done by comparison of mass spectral data with library data; Adams2 (Adams 1995) and NIST05a (NIST 2005) . In addition, structural assignments of several compounds were confirmed using authentic standards on GC/MS under the same conditions employed for the analysis of the crude volatiles. Quantification was based on calibration curves (peak area vs. concentration) generated from authentic standards of identified compounds. Experiment 4: Olfactometer Assays with a Commercial Blend Y-tube olfactometer assays (same conditions as in previous experiments) were carried out using four doses of 6.25, 12.5, 25, and 50 μl of a commercial blend comprising a 25 ng/ μl mixture of (E)-β- ocimene, (Z)-β-ocimene, and alloocimene in a 4:9:1 ratio and dissolved in 50 μl dichloromethane. Chemicals A mixture of three isomers of ocimene; (E)-βocimene, (Z)-β-ocimene, and allo-ocimene in a ratio of 4:9:1, α-pinene, hexanal, (Z)-3-hexen-1-ol, 3-carene, β-pinene, camphene, β-myrcene, linalool oxide, (E,E)-αfarnesene, α-caryophyllene, caryophyllene oxide, and βcaryophyllene were obtained from Sigma Aldrich, (St. Louis, MO, USA) while α-phellandrene, β-phellandrene, αcubebene, and α-copaene were sourced from Bedoukian, CT, USA . All the chemicals were at 95% purity. Data Analyses Using the R version 2.15.1(R Core Team 2012) statistical software, two sample t-tests were used to assess; (a) attractiveness of the individual cashew odors (time spent in the test arm) against their respective controls (time spent in the control arm), (b) frequency of sugar feeding in the presence of the three cashew test odors against the respective controls, and (c) attractiveness of the four doses of the commercial blend against the respective controls. One-way ANOVA and post-hoc analysis (Tukey’s test) also were used to compare the differences in attractiveness between the leaves, fruit, and inflorescence odor when paired to a reward and when presented alone. All tests were performed at 5% significance level.
Results Experiment 1: Olfactometer Assays with Cashew Volatiles ANOVA and Tukey’s test of the dose response assay
J Chem Ecol
Fig. 1 Dose response assay of Oecophylla longinoda to crude extracts of cashew volatiles. Bars represent ant responses (time spent as a percentage of the total time each individual ant is allocated in the olfactometer) to different doses of crude extracts of cashew volatiles. Means with asterisk are significantly different (P0.05). Means with different letters are significantly different (P
Plant Volatiles Influence the African Weaver Ant-Cashew Tree Mutualism Caroline Wanjiku & Fathiya M. Khamis & Peter E. A. Teal & Baldwyn Torto
Received: 21 May 2014 / Revised: 22 August 2014 / Accepted: 27 August 2014 # Springer Science+Business Media New York 2014
Abstract Plant volatiles influence virtually all forms of antplant symbioses. However, little is known about their role in the mutualistic relationship between the African weaver ant and the cashew tree. In this study, we tested the hypothesis that cashew tree volatiles from plant parts most vulnerable to herbivory viz. inflorescence, leaves, and fruits, are attractive to weaver ants. Using behavioral assays, we show that these volatiles attract weaver ants but without significant difference in preference for any of the odors. These same plant parts are associated with extra floral nectaries (EFNs’) and therefore we evaluated the possibility that the ants associate the volatiles with food rewards. We found that perception of the odors was followed by a searching response that led the ants to nonvolatile sugar rewards. More importantly, we observed that weaver ants spent significantly more time around the odor when it was paired to a reward. Chemical analysis of volatiles showed that the plant parts shared similarities in chemical composition, dominated by monoterpenes and sesquiterpenes. Additionally, we evaluated the attractiveness of a synthetic blend of three ocimene isomers ((E)-β-ocimene, (Z)-βocimene and allo-ocimene) identified in cashew leaf odor and shown to constitute a candidate kairomone for the cashew pest Pseudotheraptus wayi. We found that the attractiveness of the blend was dose dependent, and the response of the ants was not significantly different to that established with the C. Wanjiku : F. M. Khamis : B. Torto (*) International Centre of Insect Physiology and Ecology (icipe), P.O. Box 30772-00100, Nairobi, Kenya e-mail: [email protected] C. Wanjiku : F. M. Khamis Department of Biochemistry and Biotechnology, Kenyatta University, P.O. Box 43844, Nairobi, Kenya P. E. A. Teal USDA/ARS-Center for Medical, Agricultural and Veterinary Entomology, Gainesville, FL 32608, USA
crude volatiles from plant tissues. These results present new and interesting possibilities for improving weaver ant performance in cashew pest management. Keywords Oecophylla longinoda . Pseudotheraptus wayi . Mutualism . Olfactory cues . Plant volatiles . Cashew . Pest management . Hymenoptera . Formicidae
Introduction Ants are known natural enemies of many different insect pests, and their use dates back hundreds of years (Heil 2008). In Africa, Oecophylla longinoda (Latreille) (Hymenoptera: Formicidae), also known as the African weaver ant stands out in this realm. Having been discovered as early as 1953, this ant was and continues to be especially useful in the biocontrol of insect pests of cashew (Anacardium occidentale), coconut (Cocos nucifera), mango (Mangifera indica), and cocoa (Theobroma cacao) all across Africa (Ayenor et al. 2007; Dwomoh et al. 2008; Olotu et al. 2013; Van Mele 2008; Van Mele and Vayssières 2007; Van Mele et al. 2007; Way and Khoo 1992). Ant-plant interactions take many different forms but can be c a t e g o r i z e d b r o a d l y in t o tw o g r o u p s ; d i s p e r s a l (myrmecochory) and defense mutualisms that are either facultative or obligate (Heil 2010; Youngsteadt et al. 2009). Facultative mutualisms, are built on the exchange of food rewards; elaiosomes (specialized nutrient rich seed structures) in the case of dispersals, and food bodies or extrafloral nectaries (EFNs) in defense associations (Heil 2010; Heil and McKey 2003; Koptur and Truong 1998; Turner and Frederickson 2013). Obligate mutualisms on the other hand are based on provision of nesting sites for both dispersal and defense symbioses (Boucher et al. 1982; Heil 2008, 2010; Youngsteadt et al. 2009).
J Chem Ecol
The perennial association between the cashew tree and O. longinoda also is built on the provision of food rewards from EFNs in exchange for protection from herbivores (Rickson and Rickson 1998). The nectar secreting organs are located on panicle branches, leaves, and young fruits, which are also the most susceptible parts to herbivory (Way and Khoo 1992; Wijit 1991). The tree also accommodates honeydew producing hemipterans that are equally exploited by ants for honey dew (Way and Khoo 1992; Wijit 1991). As such, a potential defense mutualism is suggested (Rickson and Rickson 1998). Plant volatiles have been shown to mediate dispersal and defense mutualisms. In dispersals, responses of ants to seed odors (Youngsteadt et al. 2008, 2010) as well as volatiles emanating from elaiosomes have been reported (Gammans et al. 2006; Turner and Frederickson 2013). In defensive associations, ants also have been shown to respond to herbivore induced volatiles (Agrawal and Dubin-Thaler 1999; Bruna et al. 2004). At this point, the chemistry of herbivore induced damage in the cashew tree is still not clear. However, a recent study by Egonyu et al (2013), demonstrated that a blend of three components ((E)-β-ocimene, (Z)-β-ocimene, and allo-ocimene) in cashew leaf odor constitute a kairomone that elicits short-range attractive response from males of the cashew pest Pseudotheraptus wayi. This pest belongs to the order heteroptera; a group of sap sucking bugs (Aldrich 1988) ranked fourth among the most economically important groups of insects (Millar 2005). Besides cashew, P. wayi also attacks coconuts, avocados, macadamia, guava, and mango among other crops (Olotu et al. 2013). So far, the use of the predatory African weaver ant remains the most promising management strategy for the control of P. wayi (Brown 1955; Olotu et al. 2013; Vanderplank 1960), and this links small holder farmers to lucrative organic markets (Van Mele 2008). In view of this, it is important that the chemical ecology of the cashew tree system be investigated. We sought to understand how volatile organic compounds (VOCs) influence weaver ant behavior. We hypothesized that cashew volatiles emanating from the plant parts deemed most vulnerable to herbivory are attractive to the African weaver ant. To test this we carried out studies to determine: (a) the response of weaver ant workers to crude volatile extracts from cashew leaves, fruit, and inflorescence; (b) the likelihood that weaver ants relate the cashew odors to a food item, given the association of the test plant parts to EFNs; (c) the chemical composition of the volatiles and how the compounds and quantities compare to each other; and (d) the responses of the ants to a commercial blend of three isomers of ocimene, (E)-β-ocimene, (Z)-β-ocimene, and allo-ocimene that attract male Pseudotheraptus wayi. We used worker ants for this study because they play a significant role in defense and foraging (Heinze et al. 1994; Schwander et al. 2005) and in the biological control of cashew pests.
Methods and Materials Insects The weaver ant colony was established from 12 nests collected from the Mtwapa area (3.9500° S, 39.7444° E) in Kikamabala divison of Kilifi district in the coast province of Kenya in September 2012, containing approximately 200 or more ants per nest. The insects were transported to an outdoor insectary in the Animal Rearing and Containment Unit of icipe, Duduville, Nairobi campus. They were inoculated on mango seedlings (≈3½ft high) and allowed to habituate and weave fresh nests in the insectary at 24-28 °C and 55-65% relative humidity. They were maintained on a diet of 6% sugar water and powdered fish meal (silver cyprinid- “omena”) as a protein source purchased at a local store. Experiment 1: Olfactometric Assays with Cashew Volatiles A Y-tube olfactometer (arm length: 7×7 cm, internal diam: 1 cm) was used with one arm connected to the test odor; crude extract of cashew volatiles collected from leaves, fruits, and inflorescence, each dissolved in 25 μl dichloromethane (Analytical grade -Sigma Aldrich, 3050 Spruce street St. Louis, MO, USA) and loaded on a filter paper disc (3×3 cm Whatman filter paper No. 1). The other arm was connected to a control (a similar volume of solvent loaded onto a similar size and type of filter paper disc) via Teflon tubes. The crude extracts were tested in a dose response assay at 0.3, 0.6, and 1.2 plant hour equivalent (PHE) where 1 PHE=volatiles emitted by the part sampled on each individual plant per hour. An energy-saving bulb (Philips stick -11 W) emitting white light was suspended 55 cm above the bioassay arena to provide illumination. Charcoal-filtered clean air was passed through the Teflon tubes into each arm of the olfactometer at a flow rate of 348 ml/min and pulled out of the main arm of the olfactometer at the same rate by a battery powered portable vacuum pump (supplied by co-author Peter E.A. Teal; USDA/ARS-CMAVE, Gainesville, Fl, USA). Individual worker ants were released in the main arm of the Y-tube and allowed to settle for 5 min. The pump system was turned on to test the ant preferences for either the test or control odor, and each ant was given 5 min to make a choice. After every 5 insects, the odor laden filter paper and Y-tube were routinely replaced. However, when the test insect released a droplet of brown liquid from the anal region, indicating a trail laying response, the Y-tube was immediately replaced. The positions of the test and control arms of the Y-tube also were switched after every 5 insects to minimize positional bias. Preference was determined when the insect traversed the entire stem of the olfactometer and an additional 2 cm into either arm. The total time spent at each arm was recorded. Extracts of volatiles from each plant part were tested against 10 individual ants and replicated five times in each case.
J Chem Ecol
Experiment 2: Cashew Volatiles Paired with Sugar Rewards This bioassay was carried out in a glass petri-dish arena, 15 cm wide x 3.5 cm high lined with a circular wire mesh screen (mesh size≈1.19 mm) and divided into four equal quadrants. Two quadrants were designated as test and control, whereas the remaining two were left as neutral zones. The test comprised extracts of volatiles from leaves, fruit, and inflorescence (25 μl, 0.6 PHE) loaded on rubber septa. The control was rubber septum loaded with only 25 μl dichloromethane. Both septa were left to stand for about 10 min to allow the solvent to evaporate. The extract-loaded septum was suspended on the wire gauze in the test quadrant of the bioassay arena with a pin, and the other in the control quadrant. Ten droplets of glucose syrup (2:1 glucose in water) were applied on the wire mesh in both test and control quadrants using a fine camel hair brush. At this concentration, each droplet of the sugar syrup was held on the mesh screen without flow. An ant held in a small glass vial (21×34 mm) then was placed at the center of the bioassay arena and allowed to crawl out of the vial into the arena, after which the vial was removed and the arena covered with its glass cover. Each ant was given 10 min to walk around the arena, and the time spent as well as the frequency of sugar feeding in the test and control quadrant was recorded. After every 5 ants, the rubber septa were replaced and their position in the petri dish switched to minimize on positional bias. In the event of a trail laying response, the petri dish was replaced immediately before testing subsequent insects. Each odor, paired with sugar rewards, was tested against 10 individual ants and was replicated five times in each case. Experiment 3: Collection and Analysis of Cashew Volatiles Volatiles released from cashew were collected from plants growing at the Kenya Agricultural Research Institute (KARI) orchards in the Mtwapa area of Kilifi district in the coastal region of Kenya (3.9500°S and 39.7444°E). Volatiles were collected from leaves, fruit, and inflorescence on Super Q (30 mg, Analytical Research Systems, Gainesville, FL, USA) and Supelco SPME fibers (65 μm polydimethysiloxane-divinlybenzene, Sigma-Aldrich, Supelco Park, Bellefonte, PA, USA) for comparative purposes. About 12–15 leaves, 9–12 fruits, and 10–15 inflorescences were bagged in oven-baked plastic bags (355 × 508 mm, Classic Consumer products, Inc, Englewood, NJ, USA) separately, and volatiles were collected from these sources for 12 h (Super Q) and 8 h (SPME) . Each plant part was sampled four times using a different plant in each replicate. Volatiles trapped on Super Q filters were eluted using 500 μl dichloromethane (Analytical grade, Sigma Aldrich, St. Louis, MO, USA) and stored at −80 °C until use for chemical analysis and bioassays. Each individual sample contained 12 P H E . Vo l a t i l e s w e r e a n a l y z e d b y c o u p l e d g a s chromatography/mass spectrometry (GC/MS) on an Agilent
technologies Series A 7890 GC coupled to a 5975C (inert XL/ EI/CI MSD) triple axis mass detector, equipped with an HP5MS column 30 m×250 μm×0.25 μm in the electron impact mode at 70 eV. The GC oven temperature was 35 °C for 5 min with a rise of 10 °C/min to 280 ° C for 10.5 min, then 5 °C/min to 285 °C and held at this temperature for 9 min. Identification of compounds was done by comparison of mass spectral data with library data; Adams2 (Adams 1995) and NIST05a (NIST 2005) . In addition, structural assignments of several compounds were confirmed using authentic standards on GC/MS under the same conditions employed for the analysis of the crude volatiles. Quantification was based on calibration curves (peak area vs. concentration) generated from authentic standards of identified compounds. Experiment 4: Olfactometer Assays with a Commercial Blend Y-tube olfactometer assays (same conditions as in previous experiments) were carried out using four doses of 6.25, 12.5, 25, and 50 μl of a commercial blend comprising a 25 ng/ μl mixture of (E)-β- ocimene, (Z)-β-ocimene, and alloocimene in a 4:9:1 ratio and dissolved in 50 μl dichloromethane. Chemicals A mixture of three isomers of ocimene; (E)-βocimene, (Z)-β-ocimene, and allo-ocimene in a ratio of 4:9:1, α-pinene, hexanal, (Z)-3-hexen-1-ol, 3-carene, β-pinene, camphene, β-myrcene, linalool oxide, (E,E)-αfarnesene, α-caryophyllene, caryophyllene oxide, and βcaryophyllene were obtained from Sigma Aldrich, (St. Louis, MO, USA) while α-phellandrene, β-phellandrene, αcubebene, and α-copaene were sourced from Bedoukian, CT, USA . All the chemicals were at 95% purity. Data Analyses Using the R version 2.15.1(R Core Team 2012) statistical software, two sample t-tests were used to assess; (a) attractiveness of the individual cashew odors (time spent in the test arm) against their respective controls (time spent in the control arm), (b) frequency of sugar feeding in the presence of the three cashew test odors against the respective controls, and (c) attractiveness of the four doses of the commercial blend against the respective controls. One-way ANOVA and post-hoc analysis (Tukey’s test) also were used to compare the differences in attractiveness between the leaves, fruit, and inflorescence odor when paired to a reward and when presented alone. All tests were performed at 5% significance level.
Results Experiment 1: Olfactometer Assays with Cashew Volatiles ANOVA and Tukey’s test of the dose response assay
J Chem Ecol
Fig. 1 Dose response assay of Oecophylla longinoda to crude extracts of cashew volatiles. Bars represent ant responses (time spent as a percentage of the total time each individual ant is allocated in the olfactometer) to different doses of crude extracts of cashew volatiles. Means with asterisk are significantly different (P0.05). Means with different letters are significantly different (P
