- Email: [email protected]
Genetic Variability and Robustness of Host Odor Preference in Drosophila melanogaster
Current Biology 18, 1438–1443, September 23, 2008 ª2008 Elsevier Ltd All rights reserved
DOI 10.1016/j.cub.2008.08.062
Report Genetic Variability and Robustness of Host Odor Preference in Drosophila melanogaster Agnieszka Ruebenbauer,1 Fredrik Schlyter,2 Bill S. Hansson,2,3 Christer Lo¨fstedt,1 and Mattias C. Larsson2,* 1Department of Ecology Lund University So¨lvegatan 37 SE-223 62 Lund Sweden 2Department of Plant Protection Biology Division of Chemical Ecology Swedish University of Agricultural Sciences Box 102 SE-230 53 Alnarp Sweden 3Department of Evolutionary Neuroethology Max Planck Institute for Chemical Ecology Hans-Kno¨ll-Straße 8 D-07745 Jena Germany
Summary Chemosensory stimuli play a crucial role for host selection in insects, including the fruit fly Drosophila melanogaster [1]. Drosophila has been instrumental in unraveling the neurological basis of olfactory processing in insects [2]. Basic knowledge regarding chemical ecology and thorough studies of olfactory preferences are still lacking to a great extent in D. melanogaster, however. We have characterized repeatable variation in olfactory preference between five classical D. melanogaster wild-type strains toward a large array of natural host odors and synthetic compounds. By recording the rate of attraction over up to 24 hr, we could compare stimuli varying in attractiveness and characterize phenotypic parameters on the basis of individual stimuli and the whole stimulus array. Behavioral differences between strains were predominantly due to variation in a single phenotypic parameter: their overall responsiveness toward optimal and suboptimal olfactory stimuli. These differences were not explained by variation in olfactory sensitivity, locomotory activity, or general vigor monitored by survival. Comparisons with three recently established wild-type strains indicated that a high behavioral threshold against accepting suboptimal olfactory stimuli is the characteristic phenotype of wild D. melanogaster. Results and Discussion Responses of Strains to Natural and Synthetic Odors Initially, we studied behavioral responses of five classical wildtype strains, Canton-S, Oregon R-C, Oregon R-S, Berlin-K, and wild-type Berlin (Table S1 available online), to a large set of natural and synthetic odors (Figure 1). Monitoring of attraction over an extended time period demonstrated extensive
*Correspondence: [email protected]
temporal variation in olfactory-guided behavior between Drosophila wild-type strains. Much of these dynamics would probably be overlooked unless monitored over time scales that encompass the full range of responses. Most stimuli elicited a positive attraction index within 24 hr. From the temporal response patterns it is apparent, however, that the odors differ greatly in their ability to induce attraction. Natural stimuli were very attractive to the flies, but elicited significantly different responses from various strains over shorter time intervals (Figure 1A). Only banana and mango elicited more or less uniform responses from all strains; otherwise the Berlin-K (BK) and wild-type Berlin (WTB) strains consistently displayed slower responses (Figure 1A and Figure S1). Responses to synthetic attractants were overall much slower than to natural stimuli, and differences between strains were often more pronounced (Figure 1B). However, some compounds elicited slow but consistent attraction from all strains; (E)-2-hexenal was more or less neutral, whereas the two phenols, 3-octanol, and benzaldehyde were always avoided (Figure S2). WTB and BK were much more selective than the other strains in their responses to synthetic stimuli, sometimes showing only barely significant attraction even after 24 hr (Figure 1B). Lowering the concentration of repulsive odors one to two decadic steps generally changed their effect to neutral (data not shown). This dichotomy between different types of attractive and repulsive stimuli agrees relatively well with results obtained from other choice assays [3–7] but not from a recent arena assay [8]. Comparisons between starved and nonstarved animals stressed the importance of the physiological state of the flies for their willingness to respond and accentuated differences between natural odor blends and synthetic compounds as olfactory stimuli (Figure S3). In the assays, we could not quantify any aspect of fly behavior apart from trapping events. Observations of the test chambers during hourly checks indicated very few general differences in fly behavior regardless of strain and treatment, and we rarely observed apparent directedsearch behavior to any stimulus. Control Experiments: EAG, Locomotion, and Survival Electroantennographic (EAG) responses to four different synthetic stimuli (acetoin, ethyl hexanoate, ethyl acetate, and methyl salicylate) and vinegar were not different among the strains, indicating that all strains had the same ability to detect olfactory stimuli (Figure 2A). Electroantennographic recordings cannot exclude minor differences that could be revealed, e.g., by single sensillum recordings but do indicate that overall sensitivity remains the same. We also performed a series of control experiments to exclude unspecified factors that could explain differences in behavioral responses. Differences in locomotory activity could potentially affect trap catch, but a simple line-crossing assay revealed no differences in spontaneous locomotory activity (number of line crosses) between strains (means 6 SD: Canton-S 19.9 6 3.3; Oregon R-C 20.4 6 2.6; Oregon R-S 20.1 6 3.1; BK 20.3 6 2.9, WTB 20.0 6 3.2; ANOVA F = 0.2, 4 d.f., p > 0.95). Survival experiments under conditions of unlimited access to food and under starvation conditions revealed no
Genetic Variability in Host Odor Preference 1439
Figure 1. Response of Five Classical Wild-Type Strains to Selected Natural and Synthetic Stimuli Variation in attraction of five D. melanogaster classical wild-type strains over time (0–8 hr and 24 hr) to a panel of selected olfactory stimuli. (A) shows responses to selected natural stimuli, including several fruits, yeast, and balsamic vinegar. (B) shows responses to selected synthetic single compounds. Graphs show mean attraction index 6 SD. n = 5. Triangles show the time points selected in a stepwise discriminant analysis to find the greatest overall differences between strains. Responses to the full range of natural and synthetic stimuli are found in Figures S1 and S2.
apparent differences, in general vigor between strains, that could explain differences in behavioral responses (Figure 2B). What Is the Natural Behavioral Phenotype of Drosophila? Finally, we repeated behavioral experiments with the five classical strains in parallel with three newly established strains, Dalby-HL (D-HL), Helsingborg-E (HB-E), and Helsingborg-F (HB-F), to four diagnostic stimuli in order to determine which of the classical wild-type strains (if any) displayed the most ‘‘natural’’ behavioral phenotype. The newly established strains were consistently the most conservative in their choices, with behavioral responses as low as or lower than the most selective classical wild-type strains (WTB and BK) (Figure 3). Behavioral responses from the five classical wild-type strains were similar to those previously obtained from these strains (Figure 1, Table 1). Repeatability estimates (r) [9] for three of the stimuli (lemon, ethyl acetate, and acetic acid) retested with the same five strains were very high (0.98–0.99) [10]. The exception was rotten banana, which was attractive to all strains, resulting in lower between-strain variation. Behavioral differences between the classical wild-type strains probably reflect genetically determined preferences rather than differences in olfactory sensitivity, locomotion, or general vigor. Nevertheless, the most extreme differences found between wild-type strains in our study appear to be comparable in magnitude to effects caused by severe ablation
of receptor neurons [11] or to some of the variations found between different Drosophila species [12]. The picture emerging from our results is that the natural phenotype of D. melanogaster is very selective in its attraction to host odors, just like most other insects, by virtually ignoring single components of host odor blends in comparison with the complex odor blends. The relatively low selectivity displayed by the two Oregon strains and Canton-S, whose responses to some synthetic stimuli approached the levels exhibited to natural fruit odors, stands in stark contrast to the more conservative natural phenotype. Behavioral Variation and Host Choice Host and mate choice in insects are governed by highly specific chemosensory cues [13–15], with minor changes to the signal often greatly affecting behavior [14, 16–18]. Host choice is a complex event involving attraction to blends of compounds characteristic of individual or alternate hosts, in parallel with avoidance of compounds characteristic of nonhosts or unsuitable hosts [19, 20]. Chemosensory preferences in combination with physiological adaptations determine the host range [21]. Local adaptations in host preference can occur even on a microhabitat scale, causing a degree of habitat fidelity and transient, partial reproductive isolation [22]. A hostchoice event could be regarded as the product of at least two combined processes, however: evaluation of stimuli according
Current Biology Vol 18 No 18 1440
Figure 2. Control Experiments: Electroantennographic Responses and Survival in Five Strains (A) Comparisons of electroantennographic (EAG) responses from five D. melanogaster classical wild-type strains. The left shows responses to acetoin, ethyl hexanoate, ethyl acetate, methyl salicylate, and vinegar (mean net amplitude 6 SD. n = 10, pooled response from five males and five females). There were no significant differences between strains or sexes in their EAG responses to any stimulus. The right shows examples of responses to selected stimuli: water (control), acetoin, and balsamic vinegar. (B) Survival curves for five D. melanogaster classical wild-type strains (CS, OR-C, OR-S, WTB, and BK) over time (means 6 SD; n = 12; pooled response from six vials with males and six with females; each vial started with ten flies). The left panel shows survival under conditions of unlimited food supply. The right panel shows survival under starvation conditions, in which flies were provided with water only.
to a preference hierarchy, followed by a choice to accept or reject an individual host according to the degree of selectivity or choosiness [23]. This study represents the first large-scale attempt to characterize behavioral variation in this adaptive context and is thus fundamentally different from olfactory behavior in D. melanogaster on the basis of simple operational parameters such as avoidance of a single olfactory stimulus [24–26]. We have shown that strains differed primarily in their overall responsiveness toward different olfactory stimuli (Figure S4). The fruits used here all constitute acceptable oviposition sites for D. melanogaster, eliciting a high degree of attraction. Whereas banana and mango were highly attractive to all strains, other fruits were accepted to a lesser degree primarily by the Berlin strains and the newly established strains. The greatest differences between wild-type strains were found in their responses to some of the single synthetic compounds.
Responses to single compounds rarely approached those to complex blends from natural odor sources. The natural environment for the fruit fly consists of fermenting plant material rich in ethanol and other alcohols, acids, acetone, fruit esters, and highly volatile esters such as ethyl acetate. Acetoin is a metabolic product of bacteria involved in fermentation process in the presence of glucose or other fermentable source of carbon [27]. The synthetic stimuli thus constitute potentially relevant stimuli for the flies, but each represents a degraded and suboptimal signal, containing only part of the information present in the complete volatile profile of the natural stimuli (see also [28]). Preference Hierarchies and Host Shifts Factorial ANOVA with a selected time point for each stimulus always revealed strong effects of both stimulus and strain,
Genetic Variability in Host Odor Preference 1441
Figure 3. Response of Five Classical and Three New Strains to Four Diagnostic Stimuli Attraction of eight D. melanogaster wild-type strains (CS, OR-C, OR-S, WTB, BK, and the newly established D-HL, HB-E, and HB-F) over time (0–8 hr and 24 hr) to a diagnostic set of two natural and two synthetic stimuli. Graphs show mean attraction index 6 SD. n = 5. Triangles show the time points selected in a stepwise discriminant analysis to find the greatest overall differences between strains.
as well as interactions between these two factors, within all three data sets (Tables S3 and S4). Interactions suggest that strains exhibit differences in their relative attraction to olfactory stimuli. Overall preference hierarchies nevertheless appear rather robust between strains (Figure 1). Although there is obviously great variation between strains in their responsiveness to suboptimal olfactory attractants, they also seem to have retained most of their basic wild-type preferences. Other examples from mate- and host-choice systems also indicate that overall preference hierarchies and selectivity may be independently selected. Males of the cabbage looper moth Trichoplusia ni, selected to respond to a new mutant pheromone blend, responded to both the new and the old blend [29]. Host preference hierarchies of the swallowtail butterfly Papilio zelicaon remained the same in populations that used different host plants [30]. Insects in laboratory cultures can quite easily be selected to oviposit on artificial diets, or even completely artificial substrates such as paper, over just a few generations but nevertheless prefer the ancestral host ([31] and unpublished data). These examples suggest a widening of the response window rather than a change of preferences. In contrast, different host races of the apple maggot fly Rhagoletis pomonella represent a true reversal in preference, in which each host race avoids the host of the other race [32]. The combined evidence suggests that indiscriminate individuals may be favored during the initial phases of behavioral shifts, whereas a reversal in preference may require a considerably longer selection process. Genetics and Behavioral Consistency A strong genetic component determining olfactory preferences should yield strong associations between genetic similarity and behavioral responses, and perhaps also consistency of responses over different types of stimuli. Principal component analysis suggested rather high overall consistency in
clustering between different strains on the basis of their responses to different odor arrays at selected individual time points (Figure 4). Genetic differences between the strains used in this study could be caused by geographical variation [33], sampled substrate [22], random founder effects, genetic drift, and selection in captivity. Presumably, a lower degree of selectivity is caused by factors associated with captivity. The conservative behavioral phenotypes of our recently established wild-type strains (D-HL, HB-E, and HB-F) would then represent a universal behavior characteristic of wild D. melanogaster, resembled most closely by the WTB and BK strains among the classical wild-type strains. The domestication process itself does not necessarily cause low selectivity, however, given that the Berlin strains still retain a wild-type-like phenotype after many decades in captivity. Conclusions By monitoring adaptive responses over time, we have detected differences in olfactory preference among wild-type strains, which to a great extent could be attributed to variations in a single phenotypic parameter: their overall responsiveness toward optimal and suboptimal olfactory stimuli. We have demonstrated that low attraction scores to individual synthetic test stimuli in our trap assays is not a symptom of olfactory defects but rather an adaptive selectivity typical for wild D. melanogaster. Similar differences are likely to be important in any comparison between genotypes in an uncontrolled genetic background and should not be underestimated. Experimental Procedures Experimental Animals We used five different previously cultured wild-type strains (here referred to as classical wild-type strains), Canton S, Berlin-K, Oregon R-C, Oregon R-S, and wild-type Berlin (Table 1) and three recently established wild-type strains, Dalby-HL, Helsingborg-E, and Helsingborg-F. Trap Assays The trap assay is a variation on choice assays that have been used to determine differences in odor-guided behavior between different genotypes or
Table 1. Repeatability of Responses of Five Classical Strains to Four Diagnostic Stimuli Stimuli
n0
Among MSA
Within MSW
s 2A
F
Sign.
ra
Acetic acid 1% Ethyl acetate 1% Lemon Banana, rotten
2 2 2 2
0.501 1.471 0.604 0.021
0.0044 0.0021 0.0017 0.0056
0.248 0.734 0.301 0.008
114.6 692.0 355.1 3.8
*** *** *** *
0.983 0.997 0.994 0.582
a
r = s2A/(s2 + s2A), calculation according to Lessells and Boag [9]. Comparison of within-strain and between-strain variation over two separate trials (mean group size, n0) with the same stimulus, at the time point selected as most different in our stepwise discriminant analysis. MSA, mean squares among groups, MSW and s2, mean squares within groups, and s2A, among-groups variance component.
Current Biology Vol 18 No 18 1442
SPSS), in which a one-way ANOVA was used to quantify differences between the strains at each individual time point. Supplemental Data Supplemental Data include Supplemental Experimental Procedures, four figures, and four tables and can be found with this article online at http:// www.current-biology.com/cgi/content/full/18/18/1438/DC1/. Acknowledgments We are indebted to Jan-Eric Englund for invaluable contributions to the statistical analysis. Newly established fly lines were obtained from Lena Evertsson, Lie Fredholm, Roger Ha¨rdling, and A˚sa Lankinen. Fly stocks were obtained from the Bloomington Stock Center and from Bjo¨rn Brembs (WTB). Funding was provided by the Crafoord Foundation (to M.C.L.), the Trygger Foundation (to M.C.L.), Vetenskapsra˚det (The Swedish Science Council) (to C.L. and B.S.H.), and the Linnaeus Initiative ‘‘Insect Chemical Ecology, Ethology, and Evolution’’ (ICE3). Received: June 13, 2008 Revised: August 4, 2008 Accepted: August 15, 2008 Published online: September 18, 2008 References
Figure 4. Overall Similarity in Preference in Five and Eight Strains on the Basis of Multivariate Analysis Principle component analysis to separate different strains on the basis of their similarity of responses to three different sets of stimuli at selected time points. In each case, separation is based on two aggregated components, which together explained >90% of the total variation. (A) shows a comparison between five strains (CS, OR-C, OR-S, WTB, and BK) on the basis of responses to all natural stimuli. The axes have been inverted (multiplied by 21) for better spatial alignment of the results to the other two data sets. (B) shows a comparison between five strains (CS, OR-C, OR-S, WTB, and BK) on the basis of responses to synthetic stimuli. (C) shows comparison between all eight strains (CS, OR-C, OR-S, WTB, BK, and the newly established D-HL, HB-E, and HB-F) to a set of four diagnostic stimuli (two fruits and two synthetic compounds). species of Drosophila [34, 12]. Test chambers (high, plastic Petri dishes, with a ventilation hole in the lid, covered with a thin mesh) contained a treatment and control trap made from small glass vials with a cut micropipette tip inserted into the opening. The number of flies in the transparent test chamber and in either of the traps was counted at every hour during the first 8 hr and after 24 hr. Attraction to olfactory stimuli was calculated on the basis of the following attraction index, AI: AI = ðT 2 CÞ=ðT + C + OÞ where T = number of flies in the treatment trap, C = number of flies in the control trap, and O = number of flies outside the traps in the test chamber. Results of the trap assays were presented as mean attraction index of five independent repetitions 6 SD. For natural stimuli, we used yeast solution, dilute balsamic vinegar, and six different fruits: apple, banana, mango, orange, lemon, and strawberries, in both fully ripe and rotten forms. For synthetic stimuli, we selected 17 odorants (Table S2) tested at 0.1% and 1% dilutions. The odorants were chosen to represent a broad sample of ecologically relevant odors such as fruit and fermentation volatiles and odors that are known effective ligands for olfactory receptor neurons in D. melanogaster [35, 36].
Statistical Treatment of Behavioral Data For statistical analysis to draw general conclusions from the whole material, we collapsed original data sets (12,240 cases) into a series of data points consisting of a single selected time point for each stimulus, in which the strains showed the greatest difference in response. This was determined by a stepwise discriminant analysis (STEPDISC in SAS, DISCRIMINANT in
1. Schoonhoven, L.M., van Loon, J.J.A., and Dicke, M. (2005). Insect-Plant Biology (Oxford: Oxford University Press). 2. Vosshall, L.B., and Stocker, R.F. (2007). Molecular architecture of smell and taste in Drosophila. Annu. Rev. Neurosci. 30, 505–533. 3. Tully, T., and Quinn, W.G. (1985). Classical conditioning and retention in normal and mutant Drosophila melanogaster. J. Comp. Physiol. [A] 157, 263–277. 4. Charro, M.J., and Alcorta, E. (1994). Quantifying relative importance of maxillary palp information on the olfactory behavior of Drosophila melanogaster. J. Comp. Physiol. [A] 175, 761–766. 5. Devaud, J.-M. (2003). Experimental studies of adult Drosophila chemosensory behaviour. Behav. Processes. 64, 177–196. 6. Stensmyr, M.C., Dekker, T., and Hansson, B.S. (2003). Evolution of the olfactory code in the Drosophila melanogaster subgroup. Proc. Biol. Sci. 270, 2333–2340. 7. Wang, Y., Chiang, A.-S., Xia, S., Kitamoto, T., Tully, T., and Zhong, Y. (2003). Blockade of neurotransmission in Drosophila mushroom bodies impairs odor attraction, but not repulsion. Curr. Biol. 13, 1900–1904. 8. Keller, A., and Vosshall, L.B. (2007). Influence of odorant receptor repertoire on odor perception in humans and fruit flies. Proc. Natl. Acad. Sci. USA 104, 5614–5619. 9. Lessells, C.M., and Boag, P.T. (1987). Unrepeatable repeatabilitiesa common mistake. Auk 104, 116–121. 10. Magellan, K., and Magurran, A.E. (2007). Behavioural profiles: Individual consistency in male mating behaviour under varying sex ratios. Anim. Behav. 74, 1545–1550. 11. Park, S.-K., Shanbhag, S.R., Dubin, A.E., de Bruyne, M., Wang, Q., Yu, P., Shimoni, N., D’Mello, S., Carlson, J.R., Harris, G.L., et al. (2002). Inactivation of olfactory sensilla of a single morphological type differentially affects the response of Drosophila to odours. J. Neurobiol. 51, 248–260. 12. Dekker, T., Ibba, I., Siju, K.P., Stensmyr, M.C., and Hansson, B.S. (2006). Olfactory shifts parallel superspecialism for toxic fruit in Drosophila melanogaster sibling, D. sechellia. Curr. Biol. 16, 101–109. 13. Roelofs, W., Glover, T., Tang, X.H., Sreng, I., Robbins, P., Eckenrode, C., Lo¨fstedt, C., Hansson, B.S., and Bengtsson, B.O. (1987). Sex pheromone production and perception in European corn borer moths is determined by both autosomal and sex-linked genes. Proc. Natl. Acad. Sci. USA 84, 7585–7589. 14. Tasin, M., Ba¨ckman, A.-C., Coracini, M., Casado, D., Ioriatti, C., and Witzgall, P. (2006). Synergism and reduncandy in an plant volatile blend attracting grapevine moth females. Phytochemistry 68, 203–209. 15. Yang, C.H., Belawat, P., Hafen, E., Jan, L.Y., and Jan, Y.N. (2008). Drosophila egg-laying site selection as a system to study simple decision-making processes. Science 319, 1679–1683. 16. Linn, C.E., Jr., Campbell, M.G., and Roelofs, W.L. (1987). Pheromone components and active spaces: What do moths smell and where do they smell it? Science 237, 650–652.
Genetic Variability in Host Odor Preference 1443
17. Linn, C.E., Jr., Hammond, A., Du, J., and Roelofs, W.L. (1988). Specificity of male response to multicomponent pheromones in noctuid moths Trichoplusia ni and Pseudoplusia includens. J. Chem. Ecol. 14, 47–57. 18. Larsson, M.C., Hallberg, E., Kozlov, M.V., Francke, W., Hansson, B.S., and Lo¨fstedt, C. (2002). Specialized olfactory receptor neurons mediating intra- and interspecific chemical communication in leafminer moths Eriocrania spp. (Lepidoptera: Eriocraniidae). J. Exp. Biol. 205, 989–998. 19. Zhang, Q.-H., and Schlyter, F. (2003). Redundancy, synergism, and active inhibitory range of non-host volatiles in reducing pheromone attraction in European spruce bark beetle Ips typographus. Oikos 101, 299– 310. 20. Bruce, T.J.A., Wadhams, L.J., and Woodcock, C.M. (2005). Insect host location: A volatile situation. Trends Plant Sci. 10, 269–274. 21. Wiklund, C. (1975). The evolutionary relationship between adult oviposition preferences and larval host plant range in Papilio machaon L. Oecologia 18, 185–197. 22. Stalker, H.D. (1976). Chromosome studies in wild populations of D. melanogaster. Genetics 82, 323–347. 23. Reinhold, K., and Jacoby, K.J. (2002). Dissecting the repeatability of female choice in the grasshopper Chorthippus biguttulus. Anim. Behav. 64, 245–250. 24. Mackay, T.F., Hackett, J.B., Lyman, R.F., Wayne, M.L., and Anholt, R.R.H. (1996). Quantitative genetic variation of odor-guided behavior in a natural population of Drosophila melanogaster. Genetics 144, 727–735. 25. Anholt, R.R.H., Fanara, J.J., Fedorowicz, G.M., Ganguly, I., Kulkami, N.H., Mackay, T.F.C., and Rollmann, S.M. (2001). Functional genomics of odor-guided behavior in Drosophila melanogaster. Chem. Senses 26, 215–221. 26. Lavagnino, N.J., Anholt, R.R.H., and Fanara, J.J. (2008). Variation in genetic architecture of olfactory behaviour among wild-derived populations of Drosophila melanogaster. J. Evol. Biol. 10.1111/j.1420– 9101.2008.01546.x. 27. Xiao, Z.J., and Xu, P. (2007). Acetoin metabolism in bacteria. Crit. Rev. Microbiol. 33, 127–140. 28. Zhu, J., Park, K.-C., and Baker, T.C. (2003). Identification of odors from overripe mango that attract vinegar flies, Drosophila melanogaster. J. Chem. Ecol. 29, 899–909. 29. Zhu, J., Chastain, B.B., Spohn, B.G., and Haynes, K.F. (1997). Assortative mating in two pheromone strains of the cabbage looper moth, Trichoplusia ni. J. Insect Behav. 10, 805–817. 30. Wehling, W.F., and Thompson, J.N. (1997). Evolutionary conservatism of oviposition preference in a widespread polyphagous insect herbivore, Papilio zelicaon. Oecologia 111, 209–215. 31. Calatayud, P.-A., Juma, G., Njagi, P.G.N., Faure, N., Calatayud, S., Dupas, S., Le Ru, B., Magoma, G., Silvain, J.-F., and Fre´rot, B. (2008). Differences in mate acceptance and host plant recognition between wild and laboratory-reared Busseola fusca (Fuller). J. Appl. Entomol. 132, 255–264. 32. Linn, C., Feder, J., Nojima, S., Dambroski, H.R., Berlocher, S.H., and Roelofs, W. (2003). Fruit odor discrimination and sympatric host race formation in Rhagoletis. Proc. Natl. Acad. Sci. USA 100, 11490–11493. 33. Veuille, M., Baudry, E., Cobb, M., Derome, N., and Gravot, E. (2004). Historicity and the population genetics of Drosophila melanogaster and D. simulans. Genetica 120, 61–70. 34. Larsson, M.C., Domingos, A.I., Jones, W.D., Chiappe, M.E., Amrein, H., and Vosshall, L.B. (2004). Or83b encodes a broadly expressed odorant receptor essential for Drosophila olfaction. Neuron 43, 703–714. 35. Stensmyr, M.C., Giordano, E., Balloi, A., Angioy, A.-M., and Hansson, B.S. (2003). Novel natural ligands for Drosophila olfactory receptor neurones. J. Exp. Biol. 206, 715–724. 36. Hallem, E.A., and Carlson, J.R. (2006). Coding of odors by a receptor repertoire. Cell 125, 143–160.
DOI 10.1016/j.cub.2008.08.062
Report Genetic Variability and Robustness of Host Odor Preference in Drosophila melanogaster Agnieszka Ruebenbauer,1 Fredrik Schlyter,2 Bill S. Hansson,2,3 Christer Lo¨fstedt,1 and Mattias C. Larsson2,* 1Department of Ecology Lund University So¨lvegatan 37 SE-223 62 Lund Sweden 2Department of Plant Protection Biology Division of Chemical Ecology Swedish University of Agricultural Sciences Box 102 SE-230 53 Alnarp Sweden 3Department of Evolutionary Neuroethology Max Planck Institute for Chemical Ecology Hans-Kno¨ll-Straße 8 D-07745 Jena Germany
Summary Chemosensory stimuli play a crucial role for host selection in insects, including the fruit fly Drosophila melanogaster [1]. Drosophila has been instrumental in unraveling the neurological basis of olfactory processing in insects [2]. Basic knowledge regarding chemical ecology and thorough studies of olfactory preferences are still lacking to a great extent in D. melanogaster, however. We have characterized repeatable variation in olfactory preference between five classical D. melanogaster wild-type strains toward a large array of natural host odors and synthetic compounds. By recording the rate of attraction over up to 24 hr, we could compare stimuli varying in attractiveness and characterize phenotypic parameters on the basis of individual stimuli and the whole stimulus array. Behavioral differences between strains were predominantly due to variation in a single phenotypic parameter: their overall responsiveness toward optimal and suboptimal olfactory stimuli. These differences were not explained by variation in olfactory sensitivity, locomotory activity, or general vigor monitored by survival. Comparisons with three recently established wild-type strains indicated that a high behavioral threshold against accepting suboptimal olfactory stimuli is the characteristic phenotype of wild D. melanogaster. Results and Discussion Responses of Strains to Natural and Synthetic Odors Initially, we studied behavioral responses of five classical wildtype strains, Canton-S, Oregon R-C, Oregon R-S, Berlin-K, and wild-type Berlin (Table S1 available online), to a large set of natural and synthetic odors (Figure 1). Monitoring of attraction over an extended time period demonstrated extensive
*Correspondence: [email protected]
temporal variation in olfactory-guided behavior between Drosophila wild-type strains. Much of these dynamics would probably be overlooked unless monitored over time scales that encompass the full range of responses. Most stimuli elicited a positive attraction index within 24 hr. From the temporal response patterns it is apparent, however, that the odors differ greatly in their ability to induce attraction. Natural stimuli were very attractive to the flies, but elicited significantly different responses from various strains over shorter time intervals (Figure 1A). Only banana and mango elicited more or less uniform responses from all strains; otherwise the Berlin-K (BK) and wild-type Berlin (WTB) strains consistently displayed slower responses (Figure 1A and Figure S1). Responses to synthetic attractants were overall much slower than to natural stimuli, and differences between strains were often more pronounced (Figure 1B). However, some compounds elicited slow but consistent attraction from all strains; (E)-2-hexenal was more or less neutral, whereas the two phenols, 3-octanol, and benzaldehyde were always avoided (Figure S2). WTB and BK were much more selective than the other strains in their responses to synthetic stimuli, sometimes showing only barely significant attraction even after 24 hr (Figure 1B). Lowering the concentration of repulsive odors one to two decadic steps generally changed their effect to neutral (data not shown). This dichotomy between different types of attractive and repulsive stimuli agrees relatively well with results obtained from other choice assays [3–7] but not from a recent arena assay [8]. Comparisons between starved and nonstarved animals stressed the importance of the physiological state of the flies for their willingness to respond and accentuated differences between natural odor blends and synthetic compounds as olfactory stimuli (Figure S3). In the assays, we could not quantify any aspect of fly behavior apart from trapping events. Observations of the test chambers during hourly checks indicated very few general differences in fly behavior regardless of strain and treatment, and we rarely observed apparent directedsearch behavior to any stimulus. Control Experiments: EAG, Locomotion, and Survival Electroantennographic (EAG) responses to four different synthetic stimuli (acetoin, ethyl hexanoate, ethyl acetate, and methyl salicylate) and vinegar were not different among the strains, indicating that all strains had the same ability to detect olfactory stimuli (Figure 2A). Electroantennographic recordings cannot exclude minor differences that could be revealed, e.g., by single sensillum recordings but do indicate that overall sensitivity remains the same. We also performed a series of control experiments to exclude unspecified factors that could explain differences in behavioral responses. Differences in locomotory activity could potentially affect trap catch, but a simple line-crossing assay revealed no differences in spontaneous locomotory activity (number of line crosses) between strains (means 6 SD: Canton-S 19.9 6 3.3; Oregon R-C 20.4 6 2.6; Oregon R-S 20.1 6 3.1; BK 20.3 6 2.9, WTB 20.0 6 3.2; ANOVA F = 0.2, 4 d.f., p > 0.95). Survival experiments under conditions of unlimited access to food and under starvation conditions revealed no
Genetic Variability in Host Odor Preference 1439
Figure 1. Response of Five Classical Wild-Type Strains to Selected Natural and Synthetic Stimuli Variation in attraction of five D. melanogaster classical wild-type strains over time (0–8 hr and 24 hr) to a panel of selected olfactory stimuli. (A) shows responses to selected natural stimuli, including several fruits, yeast, and balsamic vinegar. (B) shows responses to selected synthetic single compounds. Graphs show mean attraction index 6 SD. n = 5. Triangles show the time points selected in a stepwise discriminant analysis to find the greatest overall differences between strains. Responses to the full range of natural and synthetic stimuli are found in Figures S1 and S2.
apparent differences, in general vigor between strains, that could explain differences in behavioral responses (Figure 2B). What Is the Natural Behavioral Phenotype of Drosophila? Finally, we repeated behavioral experiments with the five classical strains in parallel with three newly established strains, Dalby-HL (D-HL), Helsingborg-E (HB-E), and Helsingborg-F (HB-F), to four diagnostic stimuli in order to determine which of the classical wild-type strains (if any) displayed the most ‘‘natural’’ behavioral phenotype. The newly established strains were consistently the most conservative in their choices, with behavioral responses as low as or lower than the most selective classical wild-type strains (WTB and BK) (Figure 3). Behavioral responses from the five classical wild-type strains were similar to those previously obtained from these strains (Figure 1, Table 1). Repeatability estimates (r) [9] for three of the stimuli (lemon, ethyl acetate, and acetic acid) retested with the same five strains were very high (0.98–0.99) [10]. The exception was rotten banana, which was attractive to all strains, resulting in lower between-strain variation. Behavioral differences between the classical wild-type strains probably reflect genetically determined preferences rather than differences in olfactory sensitivity, locomotion, or general vigor. Nevertheless, the most extreme differences found between wild-type strains in our study appear to be comparable in magnitude to effects caused by severe ablation
of receptor neurons [11] or to some of the variations found between different Drosophila species [12]. The picture emerging from our results is that the natural phenotype of D. melanogaster is very selective in its attraction to host odors, just like most other insects, by virtually ignoring single components of host odor blends in comparison with the complex odor blends. The relatively low selectivity displayed by the two Oregon strains and Canton-S, whose responses to some synthetic stimuli approached the levels exhibited to natural fruit odors, stands in stark contrast to the more conservative natural phenotype. Behavioral Variation and Host Choice Host and mate choice in insects are governed by highly specific chemosensory cues [13–15], with minor changes to the signal often greatly affecting behavior [14, 16–18]. Host choice is a complex event involving attraction to blends of compounds characteristic of individual or alternate hosts, in parallel with avoidance of compounds characteristic of nonhosts or unsuitable hosts [19, 20]. Chemosensory preferences in combination with physiological adaptations determine the host range [21]. Local adaptations in host preference can occur even on a microhabitat scale, causing a degree of habitat fidelity and transient, partial reproductive isolation [22]. A hostchoice event could be regarded as the product of at least two combined processes, however: evaluation of stimuli according
Current Biology Vol 18 No 18 1440
Figure 2. Control Experiments: Electroantennographic Responses and Survival in Five Strains (A) Comparisons of electroantennographic (EAG) responses from five D. melanogaster classical wild-type strains. The left shows responses to acetoin, ethyl hexanoate, ethyl acetate, methyl salicylate, and vinegar (mean net amplitude 6 SD. n = 10, pooled response from five males and five females). There were no significant differences between strains or sexes in their EAG responses to any stimulus. The right shows examples of responses to selected stimuli: water (control), acetoin, and balsamic vinegar. (B) Survival curves for five D. melanogaster classical wild-type strains (CS, OR-C, OR-S, WTB, and BK) over time (means 6 SD; n = 12; pooled response from six vials with males and six with females; each vial started with ten flies). The left panel shows survival under conditions of unlimited food supply. The right panel shows survival under starvation conditions, in which flies were provided with water only.
to a preference hierarchy, followed by a choice to accept or reject an individual host according to the degree of selectivity or choosiness [23]. This study represents the first large-scale attempt to characterize behavioral variation in this adaptive context and is thus fundamentally different from olfactory behavior in D. melanogaster on the basis of simple operational parameters such as avoidance of a single olfactory stimulus [24–26]. We have shown that strains differed primarily in their overall responsiveness toward different olfactory stimuli (Figure S4). The fruits used here all constitute acceptable oviposition sites for D. melanogaster, eliciting a high degree of attraction. Whereas banana and mango were highly attractive to all strains, other fruits were accepted to a lesser degree primarily by the Berlin strains and the newly established strains. The greatest differences between wild-type strains were found in their responses to some of the single synthetic compounds.
Responses to single compounds rarely approached those to complex blends from natural odor sources. The natural environment for the fruit fly consists of fermenting plant material rich in ethanol and other alcohols, acids, acetone, fruit esters, and highly volatile esters such as ethyl acetate. Acetoin is a metabolic product of bacteria involved in fermentation process in the presence of glucose or other fermentable source of carbon [27]. The synthetic stimuli thus constitute potentially relevant stimuli for the flies, but each represents a degraded and suboptimal signal, containing only part of the information present in the complete volatile profile of the natural stimuli (see also [28]). Preference Hierarchies and Host Shifts Factorial ANOVA with a selected time point for each stimulus always revealed strong effects of both stimulus and strain,
Genetic Variability in Host Odor Preference 1441
Figure 3. Response of Five Classical and Three New Strains to Four Diagnostic Stimuli Attraction of eight D. melanogaster wild-type strains (CS, OR-C, OR-S, WTB, BK, and the newly established D-HL, HB-E, and HB-F) over time (0–8 hr and 24 hr) to a diagnostic set of two natural and two synthetic stimuli. Graphs show mean attraction index 6 SD. n = 5. Triangles show the time points selected in a stepwise discriminant analysis to find the greatest overall differences between strains.
as well as interactions between these two factors, within all three data sets (Tables S3 and S4). Interactions suggest that strains exhibit differences in their relative attraction to olfactory stimuli. Overall preference hierarchies nevertheless appear rather robust between strains (Figure 1). Although there is obviously great variation between strains in their responsiveness to suboptimal olfactory attractants, they also seem to have retained most of their basic wild-type preferences. Other examples from mate- and host-choice systems also indicate that overall preference hierarchies and selectivity may be independently selected. Males of the cabbage looper moth Trichoplusia ni, selected to respond to a new mutant pheromone blend, responded to both the new and the old blend [29]. Host preference hierarchies of the swallowtail butterfly Papilio zelicaon remained the same in populations that used different host plants [30]. Insects in laboratory cultures can quite easily be selected to oviposit on artificial diets, or even completely artificial substrates such as paper, over just a few generations but nevertheless prefer the ancestral host ([31] and unpublished data). These examples suggest a widening of the response window rather than a change of preferences. In contrast, different host races of the apple maggot fly Rhagoletis pomonella represent a true reversal in preference, in which each host race avoids the host of the other race [32]. The combined evidence suggests that indiscriminate individuals may be favored during the initial phases of behavioral shifts, whereas a reversal in preference may require a considerably longer selection process. Genetics and Behavioral Consistency A strong genetic component determining olfactory preferences should yield strong associations between genetic similarity and behavioral responses, and perhaps also consistency of responses over different types of stimuli. Principal component analysis suggested rather high overall consistency in
clustering between different strains on the basis of their responses to different odor arrays at selected individual time points (Figure 4). Genetic differences between the strains used in this study could be caused by geographical variation [33], sampled substrate [22], random founder effects, genetic drift, and selection in captivity. Presumably, a lower degree of selectivity is caused by factors associated with captivity. The conservative behavioral phenotypes of our recently established wild-type strains (D-HL, HB-E, and HB-F) would then represent a universal behavior characteristic of wild D. melanogaster, resembled most closely by the WTB and BK strains among the classical wild-type strains. The domestication process itself does not necessarily cause low selectivity, however, given that the Berlin strains still retain a wild-type-like phenotype after many decades in captivity. Conclusions By monitoring adaptive responses over time, we have detected differences in olfactory preference among wild-type strains, which to a great extent could be attributed to variations in a single phenotypic parameter: their overall responsiveness toward optimal and suboptimal olfactory stimuli. We have demonstrated that low attraction scores to individual synthetic test stimuli in our trap assays is not a symptom of olfactory defects but rather an adaptive selectivity typical for wild D. melanogaster. Similar differences are likely to be important in any comparison between genotypes in an uncontrolled genetic background and should not be underestimated. Experimental Procedures Experimental Animals We used five different previously cultured wild-type strains (here referred to as classical wild-type strains), Canton S, Berlin-K, Oregon R-C, Oregon R-S, and wild-type Berlin (Table 1) and three recently established wild-type strains, Dalby-HL, Helsingborg-E, and Helsingborg-F. Trap Assays The trap assay is a variation on choice assays that have been used to determine differences in odor-guided behavior between different genotypes or
Table 1. Repeatability of Responses of Five Classical Strains to Four Diagnostic Stimuli Stimuli
n0
Among MSA
Within MSW
s 2A
F
Sign.
ra
Acetic acid 1% Ethyl acetate 1% Lemon Banana, rotten
2 2 2 2
0.501 1.471 0.604 0.021
0.0044 0.0021 0.0017 0.0056
0.248 0.734 0.301 0.008
114.6 692.0 355.1 3.8
*** *** *** *
0.983 0.997 0.994 0.582
a
r = s2A/(s2 + s2A), calculation according to Lessells and Boag [9]. Comparison of within-strain and between-strain variation over two separate trials (mean group size, n0) with the same stimulus, at the time point selected as most different in our stepwise discriminant analysis. MSA, mean squares among groups, MSW and s2, mean squares within groups, and s2A, among-groups variance component.
Current Biology Vol 18 No 18 1442
SPSS), in which a one-way ANOVA was used to quantify differences between the strains at each individual time point. Supplemental Data Supplemental Data include Supplemental Experimental Procedures, four figures, and four tables and can be found with this article online at http:// www.current-biology.com/cgi/content/full/18/18/1438/DC1/. Acknowledgments We are indebted to Jan-Eric Englund for invaluable contributions to the statistical analysis. Newly established fly lines were obtained from Lena Evertsson, Lie Fredholm, Roger Ha¨rdling, and A˚sa Lankinen. Fly stocks were obtained from the Bloomington Stock Center and from Bjo¨rn Brembs (WTB). Funding was provided by the Crafoord Foundation (to M.C.L.), the Trygger Foundation (to M.C.L.), Vetenskapsra˚det (The Swedish Science Council) (to C.L. and B.S.H.), and the Linnaeus Initiative ‘‘Insect Chemical Ecology, Ethology, and Evolution’’ (ICE3). Received: June 13, 2008 Revised: August 4, 2008 Accepted: August 15, 2008 Published online: September 18, 2008 References
Figure 4. Overall Similarity in Preference in Five and Eight Strains on the Basis of Multivariate Analysis Principle component analysis to separate different strains on the basis of their similarity of responses to three different sets of stimuli at selected time points. In each case, separation is based on two aggregated components, which together explained >90% of the total variation. (A) shows a comparison between five strains (CS, OR-C, OR-S, WTB, and BK) on the basis of responses to all natural stimuli. The axes have been inverted (multiplied by 21) for better spatial alignment of the results to the other two data sets. (B) shows a comparison between five strains (CS, OR-C, OR-S, WTB, and BK) on the basis of responses to synthetic stimuli. (C) shows comparison between all eight strains (CS, OR-C, OR-S, WTB, BK, and the newly established D-HL, HB-E, and HB-F) to a set of four diagnostic stimuli (two fruits and two synthetic compounds). species of Drosophila [34, 12]. Test chambers (high, plastic Petri dishes, with a ventilation hole in the lid, covered with a thin mesh) contained a treatment and control trap made from small glass vials with a cut micropipette tip inserted into the opening. The number of flies in the transparent test chamber and in either of the traps was counted at every hour during the first 8 hr and after 24 hr. Attraction to olfactory stimuli was calculated on the basis of the following attraction index, AI: AI = ðT 2 CÞ=ðT + C + OÞ where T = number of flies in the treatment trap, C = number of flies in the control trap, and O = number of flies outside the traps in the test chamber. Results of the trap assays were presented as mean attraction index of five independent repetitions 6 SD. For natural stimuli, we used yeast solution, dilute balsamic vinegar, and six different fruits: apple, banana, mango, orange, lemon, and strawberries, in both fully ripe and rotten forms. For synthetic stimuli, we selected 17 odorants (Table S2) tested at 0.1% and 1% dilutions. The odorants were chosen to represent a broad sample of ecologically relevant odors such as fruit and fermentation volatiles and odors that are known effective ligands for olfactory receptor neurons in D. melanogaster [35, 36].
Statistical Treatment of Behavioral Data For statistical analysis to draw general conclusions from the whole material, we collapsed original data sets (12,240 cases) into a series of data points consisting of a single selected time point for each stimulus, in which the strains showed the greatest difference in response. This was determined by a stepwise discriminant analysis (STEPDISC in SAS, DISCRIMINANT in
1. Schoonhoven, L.M., van Loon, J.J.A., and Dicke, M. (2005). Insect-Plant Biology (Oxford: Oxford University Press). 2. Vosshall, L.B., and Stocker, R.F. (2007). Molecular architecture of smell and taste in Drosophila. Annu. Rev. Neurosci. 30, 505–533. 3. Tully, T., and Quinn, W.G. (1985). Classical conditioning and retention in normal and mutant Drosophila melanogaster. J. Comp. Physiol. [A] 157, 263–277. 4. Charro, M.J., and Alcorta, E. (1994). Quantifying relative importance of maxillary palp information on the olfactory behavior of Drosophila melanogaster. J. Comp. Physiol. [A] 175, 761–766. 5. Devaud, J.-M. (2003). Experimental studies of adult Drosophila chemosensory behaviour. Behav. Processes. 64, 177–196. 6. Stensmyr, M.C., Dekker, T., and Hansson, B.S. (2003). Evolution of the olfactory code in the Drosophila melanogaster subgroup. Proc. Biol. Sci. 270, 2333–2340. 7. Wang, Y., Chiang, A.-S., Xia, S., Kitamoto, T., Tully, T., and Zhong, Y. (2003). Blockade of neurotransmission in Drosophila mushroom bodies impairs odor attraction, but not repulsion. Curr. Biol. 13, 1900–1904. 8. Keller, A., and Vosshall, L.B. (2007). Influence of odorant receptor repertoire on odor perception in humans and fruit flies. Proc. Natl. Acad. Sci. USA 104, 5614–5619. 9. Lessells, C.M., and Boag, P.T. (1987). Unrepeatable repeatabilitiesa common mistake. Auk 104, 116–121. 10. Magellan, K., and Magurran, A.E. (2007). Behavioural profiles: Individual consistency in male mating behaviour under varying sex ratios. Anim. Behav. 74, 1545–1550. 11. Park, S.-K., Shanbhag, S.R., Dubin, A.E., de Bruyne, M., Wang, Q., Yu, P., Shimoni, N., D’Mello, S., Carlson, J.R., Harris, G.L., et al. (2002). Inactivation of olfactory sensilla of a single morphological type differentially affects the response of Drosophila to odours. J. Neurobiol. 51, 248–260. 12. Dekker, T., Ibba, I., Siju, K.P., Stensmyr, M.C., and Hansson, B.S. (2006). Olfactory shifts parallel superspecialism for toxic fruit in Drosophila melanogaster sibling, D. sechellia. Curr. Biol. 16, 101–109. 13. Roelofs, W., Glover, T., Tang, X.H., Sreng, I., Robbins, P., Eckenrode, C., Lo¨fstedt, C., Hansson, B.S., and Bengtsson, B.O. (1987). Sex pheromone production and perception in European corn borer moths is determined by both autosomal and sex-linked genes. Proc. Natl. Acad. Sci. USA 84, 7585–7589. 14. Tasin, M., Ba¨ckman, A.-C., Coracini, M., Casado, D., Ioriatti, C., and Witzgall, P. (2006). Synergism and reduncandy in an plant volatile blend attracting grapevine moth females. Phytochemistry 68, 203–209. 15. Yang, C.H., Belawat, P., Hafen, E., Jan, L.Y., and Jan, Y.N. (2008). Drosophila egg-laying site selection as a system to study simple decision-making processes. Science 319, 1679–1683. 16. Linn, C.E., Jr., Campbell, M.G., and Roelofs, W.L. (1987). Pheromone components and active spaces: What do moths smell and where do they smell it? Science 237, 650–652.
Genetic Variability in Host Odor Preference 1443
17. Linn, C.E., Jr., Hammond, A., Du, J., and Roelofs, W.L. (1988). Specificity of male response to multicomponent pheromones in noctuid moths Trichoplusia ni and Pseudoplusia includens. J. Chem. Ecol. 14, 47–57. 18. Larsson, M.C., Hallberg, E., Kozlov, M.V., Francke, W., Hansson, B.S., and Lo¨fstedt, C. (2002). Specialized olfactory receptor neurons mediating intra- and interspecific chemical communication in leafminer moths Eriocrania spp. (Lepidoptera: Eriocraniidae). J. Exp. Biol. 205, 989–998. 19. Zhang, Q.-H., and Schlyter, F. (2003). Redundancy, synergism, and active inhibitory range of non-host volatiles in reducing pheromone attraction in European spruce bark beetle Ips typographus. Oikos 101, 299– 310. 20. Bruce, T.J.A., Wadhams, L.J., and Woodcock, C.M. (2005). Insect host location: A volatile situation. Trends Plant Sci. 10, 269–274. 21. Wiklund, C. (1975). The evolutionary relationship between adult oviposition preferences and larval host plant range in Papilio machaon L. Oecologia 18, 185–197. 22. Stalker, H.D. (1976). Chromosome studies in wild populations of D. melanogaster. Genetics 82, 323–347. 23. Reinhold, K., and Jacoby, K.J. (2002). Dissecting the repeatability of female choice in the grasshopper Chorthippus biguttulus. Anim. Behav. 64, 245–250. 24. Mackay, T.F., Hackett, J.B., Lyman, R.F., Wayne, M.L., and Anholt, R.R.H. (1996). Quantitative genetic variation of odor-guided behavior in a natural population of Drosophila melanogaster. Genetics 144, 727–735. 25. Anholt, R.R.H., Fanara, J.J., Fedorowicz, G.M., Ganguly, I., Kulkami, N.H., Mackay, T.F.C., and Rollmann, S.M. (2001). Functional genomics of odor-guided behavior in Drosophila melanogaster. Chem. Senses 26, 215–221. 26. Lavagnino, N.J., Anholt, R.R.H., and Fanara, J.J. (2008). Variation in genetic architecture of olfactory behaviour among wild-derived populations of Drosophila melanogaster. J. Evol. Biol. 10.1111/j.1420– 9101.2008.01546.x. 27. Xiao, Z.J., and Xu, P. (2007). Acetoin metabolism in bacteria. Crit. Rev. Microbiol. 33, 127–140. 28. Zhu, J., Park, K.-C., and Baker, T.C. (2003). Identification of odors from overripe mango that attract vinegar flies, Drosophila melanogaster. J. Chem. Ecol. 29, 899–909. 29. Zhu, J., Chastain, B.B., Spohn, B.G., and Haynes, K.F. (1997). Assortative mating in two pheromone strains of the cabbage looper moth, Trichoplusia ni. J. Insect Behav. 10, 805–817. 30. Wehling, W.F., and Thompson, J.N. (1997). Evolutionary conservatism of oviposition preference in a widespread polyphagous insect herbivore, Papilio zelicaon. Oecologia 111, 209–215. 31. Calatayud, P.-A., Juma, G., Njagi, P.G.N., Faure, N., Calatayud, S., Dupas, S., Le Ru, B., Magoma, G., Silvain, J.-F., and Fre´rot, B. (2008). Differences in mate acceptance and host plant recognition between wild and laboratory-reared Busseola fusca (Fuller). J. Appl. Entomol. 132, 255–264. 32. Linn, C., Feder, J., Nojima, S., Dambroski, H.R., Berlocher, S.H., and Roelofs, W. (2003). Fruit odor discrimination and sympatric host race formation in Rhagoletis. Proc. Natl. Acad. Sci. USA 100, 11490–11493. 33. Veuille, M., Baudry, E., Cobb, M., Derome, N., and Gravot, E. (2004). Historicity and the population genetics of Drosophila melanogaster and D. simulans. Genetica 120, 61–70. 34. Larsson, M.C., Domingos, A.I., Jones, W.D., Chiappe, M.E., Amrein, H., and Vosshall, L.B. (2004). Or83b encodes a broadly expressed odorant receptor essential for Drosophila olfaction. Neuron 43, 703–714. 35. Stensmyr, M.C., Giordano, E., Balloi, A., Angioy, A.-M., and Hansson, B.S. (2003). Novel natural ligands for Drosophila olfactory receptor neurones. J. Exp. Biol. 206, 715–724. 36. Hallem, E.A., and Carlson, J.R. (2006). Coding of odors by a receptor repertoire. Cell 125, 143–160.