ARTICLE IN PRESS

Journal of Insect Physiology 53 (2007) 509516 www.elsevier.com/locate/jinsphys

Chromatic cues to trap the oriental fruit y, Bactrocera dorsalis

Wen-Yen Wua, Yu-Po Chena, En-Cheng Yanga,b,
b

Department of Entomology, National Chung Hsing University, Taichung, Taiwan Laboratory of Insect Neurobiology, Department of Entomology, National Taiwan University, Taipei 10617, Taiwan Received 27 December 2005; received in revised form 5 February 2007; accepted 5 February 2007

Abstract Various colors have been used as visual cues to trap insect pests. For example, yellow traps for monitoring and control of the oriental fruit y (Bactrocera dorsalis) have been in use for a very long time. However, the chromatic cue of using color traps has never been meticulously investigated. In this study, the spectral sensitivities of the photoreceptors in the compound eyes of B. dorsalis were measured intracellularly, and the theory of receptor quantum catch was applied to study the chromatic cue of y attracting. Responses to ve wavelength categories with peak wavelengths of 370, 380, 490, and 510 nm, and one with dual peaks at 350 and 490 nm were recorded. Based on spectral sensitivities, six colored papers were chosen to test the color preference of the y, and an additional UV preference test was done to conrm the effect of the UV stimuli. It was concluded that UV and green stimuli (spectra: 300380 nm and 500570 nm) would enhance the attractiveness of a colored paper to the oriental fruit y, and blue stimuli (380500 nm) would diminish the attractiveness. r 2007 Elsevier Ltd. All rights reserved.
Keywords: Spectral sensitivity; Color preference; Oriental fruit y; Bactrocera dorsalis

1. Introduction Several species of fruit ies (Tephritidae) are major and widespread destructive pests in agriculture around the world (Christenson and Foote, 1960). Their oviposition causes serious damage to fruit crops, costing millions of dollars every year. To control and monitor these pests, many studies have been devoted to trap-designs (Katsoyannos, 1989; Roessler, 1989; Heath et al., 1993). Since host-nding is the indispensable stage for discovering a suitable place for oviposition, and since it is strongly inuenced by both visual and olfactory cues (Prokopy and Owens, 1983; Prokey et al., 1990), the best strategy to attract these ies is by using these cues to develop effective traps to lure them. The oriental fruit y, Bactrocera dorsalis, is the main fruit y pest in the Pacic Rim (Haramoto and Bess, 1970;
Corresponding author. Laboratory of Insect Neurobiology, Department of Entomology, National Taiwan University, Taipei 10617, Taiwan. Tel.: +886 2 33669640; fax: +886 2 33652092. E-mail address: ecyang@ntu.edu.tw (E.-C. Yang).

Liu, 1981), and numerous lure-and-kill traps with yellow colored surfaces have been developed and applied in the eld, such as yellow sticky papers and the famous malespecic methyl eugenol-baited traps (Cornelius et al., 1999; Alyokhin et al., 2000; Chen and Dong, 2001). Both Vargas et al. (1991) and Cornelius et al. (1999) demonstrated that yellow fruit-mimicking spheres were an excellent device to lure the oriental fruit y, and Alyokhin et al. (2000) demonstrated that the visual cues of a yellow fruitmimicking sphere trap could increase the attractiveness of hydrolyzed liquid protein odor. It has been shown that visual and olfactory cues can function separately and have a complementary effect to each other for attracting ies (Jang and Light, 1991; Vargas et al., 1991; Cornelius et al., 1999; Alyokhin et al., 2000; Corneluis et al., 2000a, b). Thus, to optimize lure-and-kill traps, these cues must be investigated carefully using experiments based on neuroethological aspects. However, most of these studies were done only on olfactory cues for tephritids (Light et al., 1992; Raptopoulos et al., 1995). Also, it should be noted that using a color name as a reference to describe a ys preference is not correct, since our color vision is quite

0022-1910/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.jinsphys.2007.02.003

ARTICLE IN PRESS
510 W.-Y. Wu et al. / Journal of Insect Physiology 53 (2007) 509516

different from that of an insect, and it is better to dene the ys chromatic preference using the insect visible spectrum. In this paper, we report the wavelength preference of the oriental fruit y based on the evidence from electrophysiological recordings complete with behavioral demonstration. Since the host-nding behavior is closely related to matending and oviposition behaviors in fruit ies (Prokopy and Owens, 1983), it can be assumed that the inherent wavelength preference is also closely related to host-nding behavior. From a physiological perspective, the chromatic cues of host-nding behavior can be taken as a different proportion of stimulation for each type of photoreceptor. Thus, selecting the chromatic stimuli according to the spectral sensitivity of the photoreceptors for testing the ys wavelength preference will yield more specic results than studies referring to the host fruit of the y only (Vargas et al., 1991; Cornelius et al., 1999). We tested the ys preference for colored papers that were selected based on the spectral sensitivities of a ys photoreceptors, and analyzed the results to reveal possible chromatic cues. 2. Materials and methods 2.1. Animals Experiments were carried out on both male and female adult oriental fruit ies, B. dorsalis, derived from a laboratory stock maintained at 2871 1C in a rearing room with 12:12 (L:D) photoperiod and fed with the peptone sugar mixture. 2.2. Spectral sensitivity measurement Before electrophysiological manipulation, the experimental y was immobilized by placing it in a freezer with chipped ice at 4 1C for 1530 min. Then the head of the y was mounted on a brass pedestal, with a beeswax/rosin (3:1) mixture, so that the head and thorax were rigidly secured, and the abdomen was free to perform ventilatory movements. A small window was cut in the dorsallateral part of the left compound eye to expose the retinula for inserting the recording microelectrode, and a silver wire was inserted into the thorax as the indifferent electrode. The damaged surfaces were covered immediately with vaseline to prevent drying. The brass pedestal with the mounted animal was placed on a metal plate with the animals head at the center of a Cardan arm perimeter, which was mounted with an optical ber-guided stimulating light source. The microelectrode was made of microlament aluminosilicate capillary glass (O.D. 1.0 mm, I.D. 0.68 mm, AF100-68-10, Sutter Instrument Co.) pulled on a FlamingBrown microelectrode puller (P-97 Flaming/Brown Micropipette Puller, Sutter Instrument Co.), and had a resistance of 140160 MO when lled with 1 M lithium chloride solution. Using a micromanipulator (MWS-32, Narishige

Scientic Instrument Lab.), the microelectrode was adjusted and lowered vertically to insert into the retinula through the small window in the compound eye. After inserting the microelectrode, the y was dark-adapted for at least 10 min. The following manipulations were performed in the dark room to keep the y dark-adapted. The microelectrode was advanced meticulously, and a gentle tapping was performed when the tip of microelectrode was going to penetrating through the cell membrane. Each time the baseline potential shifted, a series of ash stimuli were used to check if the tip of microelectrode was impaling the photoreceptor. If a depolarization response was detected, the Cardan arm perimeter would be adjusted to align with the visual axis of the recorded cell by obtaining the maximum response. The recorded cell would be further veried as a photoreceptor that the cell had a depolarized graded response waveform, which was composed of an initial on-transient depolarizing peak and a sustained plateau, to an on-axis saturated white light stimulus of 200 ms duration. The signals measured by the microelectrode were preamplied 10 by an amplier (Neuropobe, Model 1600, A-M Systems Inc.) and then sent via a 12-bit multifunction data acquisition system (PCI-6024E, National Instruments) to an IBM compatible PC. A Xenon-short arc lamp (XBO 1000W/HS/OFR, OSRAM) was used as the stimulating light source. The polychromatic white light was guided to pass a quartz circular variable neutral-density wedge lter (Acton Research Co.), which was capable of varying the intensity of the light over a range of approximately 3 log units, and then sent into the monochromator (SP-150-M with 150030-300 grating, Acton Research Co.). The separated monochromatic light with half-band widths below 10 nm was controlled by a magnetic shutter (SH-150, Acton Research Co.) to form the ash stimulation with 20 ms duration and was then guided by a UV-VIS ber optic bundle (LG-455-020-3, Acton Research Co.) to project on the recorded eye. Since the terminal of the ber optic bundle with a diameter of approximately 1.5 mm was 230 mm away from the recorded eye, the terminal of the ber optic bundle therefore subtended an angle less than 0.371 at the eye. The optic instruments and the multifunction data acquisition system were controlled by a program developed by LabVIEW software (ver. 6i, National Instruments) and executed on the PC. Thus, it could produce equal quanta of ux stimuli at 41 wavelengths from 300 to 700 nm, with 10 nm steps, and record the respective electrophysiological responses from the photoreceptor. The amplitude of the recorded responses were measured on-line and calculated as the spectral sensitivity by the same program. All the measured spectral sensitivity curves were calculated as the mean of squared errors with the theoretical visual pigment absorption curves, which were derived from the nomogram by Ebrey and Honig (1977), at 41 peak wavelengths from 300 to 700 nm. The lmax was determined at the wavelength with the minimum mean square error.

ARTICLE IN PRESS
W.-Y. Wu et al. / Journal of Insect Physiology 53 (2007) 509516 511

2.3. Color preference test To test the color preference of the oriental fruit y, 6 colored papers (black, white, green, yellow, orange and red) were chosen according to the spectral sensitivities of the photoreceptors. Reectance spectra of these colored papers were obtained with a spectrometer (S2000, Ocean Optics Inc.) compared with a reectance standard (Spectralon diffuse white standard, Labsphere, Inc.). The black colored paper with almost no reectance was chosen as the test control. Yellow, orange and red colored papers were chosen to test the effect of different stimuli on the sensitivity peaks between 500 and 560 nm. The green and white colored papers were chosen to test the effect of strong and no UV stimulus (Fig. 1).

120 100 Reflectance (%) 80 60 40 20 0 300 400 500 Wavelength (nm) 600 700

Experiments were conducted on sunny days from April to June 2004 on National Chung Hsing University campus. Tests were performed in a dodecagon maze (Fig. 2) outdoors, with two sets of colored papers attached symmetrically on the side walls in random orders and with each colored paper separated with opaque black plastic plating. To allow the ies to see all 6 colored papers in any horizontal direction, the tested ies restricted in a transparent transport container were put at the center of the maze before they were set free. After 35 min, the container was opened to set the les free and allowed them to move and y freely to lower any possible stress effect. The number of ies in each inside-chamber was counted after 30 min. During the tests, the dodecagon maze with a transparent top was placed in the shade of trees in order to prevent any possible effects of polarization from incident light. Since it has been reported that the ERG of the y may be subject to age-related changes (Loew, 1975), and that visual attractive cues for tephritid ies may differ between sexes (Sivinski, 1990), the ies were tested separately for different ages and sex with three replicates. Both male and female ies from 1 to 20 days old from eclosion were tested. Each test was performed on at least 30 ies; no y was tested more than once. 2.4. UV preference test According to the result of color preference test (see below and Fig. 8), the white colored paper with a high UV reectance was less attractive to the oriental fruit y than the green colored paper and even less attractive than the black colored paper, a further test was performed to evaluate the affect of UV stimuli on the color preference. Twelve white colored papers were attached onto the inside walls of the dodecagon maze as the above description. On

Fig. 1. Reectance spectra of colored papers used in the color preference test.

Fig. 2. Sketch of the dodecagon maze with 12 inside-chambers used for simultaneously comparing the attractiveness of different colored papers to oriental fruit ies. The numbers indicate the dimensions of the maze and inside-chambers in centimeters. The structures of the inside-chambers are identically, and the right illustration indicates the dimension of one of the inside-chambers marked in dark-gray color.

ARTICLE IN PRESS
512 W.-Y. Wu et al. / Journal of Insect Physiology 53 (2007) 509516

the top of the six chambers of the maze, UV-cut transparent plastic plates servicing as UV cut-off lters were applied to produce white colored paper without UV reection below 380 nm, but the plates still had more than 80% transmission within 390700 nm. The other six chambers without the transparent plastic plates were illuminated by normal light to the white colored papers, and thus providing higher UV reection from the papers. All the six UV-cut chambers were separated by the normal illuminated chambers. Therefore, the setup could provide dual-choice of white colored papers with and without UV illumination below 380 nm. The experimental procedure followed the color preference test. Because the color preference was not signicantly different between the sexes and the different ages (see color preference test result), the tests were performed with more than 25 males and 25 females with the age of 7 days old together at the same time. Each test was performed with at least 50 ies; ies were tested only once. 3. Results 3.1. Spectral sensitivities of the compound eye photoreceptors of the oriental fruit y Responses to ve wavelength categories were measured from 37 cells in the dorsalfrontal part of the left compound eye of the oriental fruit y in dark adaptation. The maximum intracellular response of the oriental fruit y photoreceptors was generally up to 70 mV. In some recordings, however, such as in the 380 nm and the 510 nm photoreceptors, the recorded maximum response could only reach about 2540 mV. All recordings were validated as intracellular by means of characteristics such as the baseline drop during penetration of the cell membrane (35 to 60 mV) and the transient-tonic depolarized graded response waveform to a sufcient light stimulus. The low maximum responses of these recordings could have been caused by the ltering effects of the special histological superposition condition (Wu et al., 1985), or more likely the result of a microelectrode location distant from the cell body of the photoreceptor. According to their lmax of the spectral sensitivity curve, these photoreceptors were grouped and termed as 370, 380, 490, 510 nm and dual-peak photoreceptors. Further details of these photoreceptors are described as follows: 370 nm photoreceptor: A single cell, with maximum spectral sensitivity at 370 nm, was measured with three measurements, one bi-directional run (300700 nm and 700300 nm) and another uni-directional run (from 300 to 700 nm). The black solid line of Fig. 3 shows that the sensitive range of this cell was mainly located in the UV range of 300400 nm. The tting procedure also demonstrated that the curve was best tted by the theoretical visual pigment absorption curve of 370 nm visual pigment. 380 nm photoreceptor: The spectral sensitivity curve of the 380 nm photoreceptor was similar to that of the 370 nm

100 Relative Sensitivity (%)

75

50

25

0 300 400 500 Wavelength (nm)

Fig. 3. The relative spectral sensitivity curve of the oriental fruit y 370 nm photoreceptor (data averaged from three measurements of cell 15, the scatter bar gives the standard error). The light gray dash line shows the theoretical visual pigment absorption curve with the peak wavelength at 370 nm. The gray dot line shows the spectral sensitivity curve of the R7p photoreceptor in the dipteran retinula (Hardie, 1986).

600

700

100

Relative Sensitivity (%)

75

50

25

0 300 400 500 Wavelength (nm) 600 700

Fig. 4. The relative spectral sensitivity curve of the oriental fruit y 380 nm photoreceptor (data averaged from two measurements of cell 23). The light gray dash line shows the theoretical visual pigment absorption curve with the peak wavelength at 380 nm. The gray dot line shows the spectral sensitivity curve of the R7y photoreceptor in the dipteran retinula (Hardie, 1986).

photoreceptor in the UV range (Fig. 4), and formed a plateau shoulder at 440520 nm with sensitivity of about 40% of the peak. The peak of the sensitivity curve was at 370 nm, but due to the fact that the averaged curve is best tted with the theoretical visual pigment absorption curve of 380 nm visual pigment (Fig. 4), this type is therefore referred to as the 380 nm photoreceptor. The cell was measured successfully with two measurements (bi-directional run). 490 nm photoreceptor: Three measurements were successfully made on the 490 nm photoreceptor (one bi-directional and another uni-directional run). This photoreceptor

ARTICLE IN PRESS
W.-Y. Wu et al. / Journal of Insect Physiology 53 (2007) 509516 513

possesses a simple spectral sensitivity peak at 490 nm (Fig. 5), and the spectral sensitivity curve is best tted with the theoretical visual pigment absorption curve of 490 nm visual pigment. 510 nm photoreceptor: The shape of the spectral sensitivity curve of the 510 nm photoreceptor appeared to be like a screened curve of the 490 nm photoreceptor (Fig. 6). Therefore, the spectral sensitivity peak was shifted to the range of 480550 nm, and the curve was best tted with the theoretical visual pigment absorption curve of 510 nm visual pigment (Fig. 6). The cell was also measured with

three repeats (one bi-directional and another uni-directional run). Dual-peak photoreceptor: These photoreceptors were the photoreceptor most often encountered in 33 cells in the retinula. The spectral sensitivity curves of these cells were composed of two major peaks at the UV (330370 nm) and the blue-green (430510 nm) range. The band widths of the peaks and the amplitude ratios between the peaks showed some varieties. Fig. 7 shows the spectral sensitivity curve averaged from the measurement of 7 cells, the lmax was obtained at 350 and 490 nm by tting with the theoretical visual pigment absorption curve. 3.2. Color preference of the oriental fruit y The preferred choice rates were obtained from three replicates of each age and sex. These rates were not signicantly related to either age or sex when analyzed with a three-way ANOVA test, but they were signicantly related to the color of the colored papers (F 88.77; df 5; Po0.0001; ANOVA Procedure of SASs). Choice rates were also not signicantly related to the mature status of the ies (17 days old ies and older ies) when analyzed with a two-way ANOVA test, but were signicantly related to the color of the colored papers (F 73.88; df 5; Po0.0001; ANOVA Procedure of SASs). Therefore the data from both sexes and all ages were combined for further analyses. Fig. 8 shows the mean choice rates for the colored papers from 120 replicates (three replicates for 20 ages and two sexes). The green, yellow and orange colored papers, with an oriental fruit y attracting rate of (Mean7SE) 0.2265270.00766, 0.2194170.00826, and 0.1913270.00704, respectively, were signicantly more attractive than the red, black, or white colored papers
100

100

Relative Sensitivity (%)

75

50

25

0 300 400 500 Wavelength (nm) 600 700

Fig. 5. The relative spectral sensitivity curve of the oriental fruit y 490 nm photoreceptor (data averaged from three measurements of cell 40, the scatter bar gives the standard error). The light gray dash line shows the theoretical visual pigment absorption curve with the peak wavelength at 490 nm. The gray dot line shows the spectral sensitivity curve of the R8p photoreceptor in the dipteran retinula (Hardie, 1986).

100

Relative Sensitivity (%)

Relative Sensitivity (%)

75

75

50

50

25

25

0 300 400 500 Wavelength (nm) 600 700

0 300 400 500 Wavelength (nm) 600 700

Fig. 6. The relative spectral sensitivity curve of the oriental fruit y 510 nm photoreceptor (data averaged from six measurements of cell 14, the scatter bar gives the standard error). The light gray dash line shows the theoretical visual pigment absorption curve with the peak wavelength at 510 nm. The gray dot line shows the spectral sensitivity curve of the R8y photoreceptor in the dipteran retinula (Hardie, 1986).

Fig. 7. The relative spectral sensitivity curve of the oriental fruit y dualpeak photoreceptors (data averaged from seven cells, the scatter bar gives the standard error). The light gray dash lines show the theoretical visual pigment absorption curves with the peak wavelength at 350 and 490 nm. The gray dot line shows the spectral sensitivity curve of the R16 photoreceptors in the dipteran retinula (Hardie, 1986).

ARTICLE IN PRESS
514 W.-Y. Wu et al. / Journal of Insect Physiology 53 (2007) 509516

0.25

a,b b c

0.20 Attraction Rate

0.15

0.10

0.05

0.00 Yellow Green Orange Red Color Paper Black White

Fig. 8. Attraction (Mean+SE) of different colored papers to the oriental fruit y. The same lowercase letters above the bars in (ac) indicate no signicant difference (F 73.61; df 5; Po0.0001; ANOVA Procedure and Tukeys Studentized Range (HSD) Test of SASs).

(Fig. 8). The white colored paper with the 0.079970.00504 oriental fruit y attracting rate was signicantly less attractive to the oriental fruit y than the red and black colored paper with an attracting rate of 0.1487470.00707 and 0.1341170.00578, respectively (Fig. 8; F 73.61; df 5; Po0.0001; ANOVA Procedure of SASs). 3.3. UV preference test The choice rates were obtained from 3 replicates, which showed that the white colored papers with the UV illumination below 380 nm were signicantly more attractive than the white colored papers without the UV illumination (Mean7SE: 0.4670.03055 vs. 0.246677 0.04807; F 13.17; df 1; Po0.05; ANOVA Procedure and Tukeys Studentized Range (HSD) Test of SASs). 4. Discussion 4.1. Spectral sensitivity of the photoreceptors Dual-peak photoreceptor: The dual-peak photoreceptors have a similar action spectrum to the R16 photoreceptors found in other dipteran ies (Fig. 7; Horridge and Mimura, 1975; Horridge et al., 1975; Hardie, 1979, 1986; Hardie et al., 1979). The antenna pigment hypothesis proposed by Kirschfeld et al. (1977) suggested that the blue-green peak was from the intrinsic blue-green sensitive rhodopsins and that the UV peak was due to the UV sensitizing pigments transferring their absorbed energy to the rhodopsins. This would then make the action spectra of these photoreceptors become broadly distributed with two peak wavelengths. Variation of the spectral sensitivity curves can be assumed to be the result of self-absorption or electric coupling effects (Manzel, 1975).

370 and 380 nm photoreceptors: The spectral sensitivity curve of the 370 nm photoreceptor is similar to the curve of the R7p photoreceptor obtained from other ies (Fig. 2; Horridge et al., 1975; Hardie, 1979, 1986; Hardie et al., 1979). It is very likely that the sensitivity range of the 370 nm photoreceptor may be elicited by a single visual pigment with its lmax at 370 nm. In contrast, the 380 nm photoreceptor has a more complicated spectral sensitivity curve, in which a shoulder was observed from 440 to 520 nm. This curve is also similar to the R7y photoreceptor of other dipteran ies (Fig. 3; Horridge et al., 1975; Hardie et al., 1979; Hardie, 1986). According to Hardie (1986), the mechanism that elicits these action spectra of the photoreceptors may possibly be caused by the fact that the blue sensitive rhodopsins (e.g. xanthopsin, lmax 430 nm) of the cell were greatly hindered by the blue screening pigment (e.g. carotenoid). However, the cells possess a UV sensitizing pigment similar to the dual-peak photoreceptors and the absorbed energy of the unscreened sensitizing pigments can transfer to the blue sensitive rhodopsins and cause the UV peak with the blue shoulder in the action spectra of these cells. Therefore, the 380 nm photoreceptor has a complicated action spectrum, and the summation of UV sensitizing pigment and the hindered blue sensitive rhodopsins makes the spectral sensitivity curve best t with the 380 nm theoretical visual pigment absorption curve. In addition, the block effects of the blue screening pigments in the 380 nm photoreceptor of the oriental fruit y would be less than those in the R7y photoreceptor of other dipteran ies, as the 380 nm photoreceptor was more sensitive than the R7y photoreceptor within the range between 420 and 510 nm. 490 and 510 nm photoreceptors: The spectral sensitivity curves of 490 and 510 nm photoreceptors are similar to the R8p and R8y photoreceptors of other dipteran ies (Figs. 4 and 5; Hardie et al., 1979; Hardie, 1986). It is possible that the 490 and 510 nm photoreceptors of oriental fruit y are the R8 cells described by Wu et al. (1985). Because the rhabdomere of the R8 cell is beneath the rhabdomere of the R7 cell in the oriental fruit y (Wu et al., 1985), the 370 and 380 nm photoreceptors may be considered as the R7 cells. If this is the case, the spectral reception of the 490 and 510 nm photoreceptors should be ltered by the 370 and 380 nm photoreceptors, respectively. Because the spectral absorptions of 490 and 370 nm photoreceptors are not overlapping in their dominant peaks, the 490 nm photoreceptor still keeps its spectral sensitivity in the blue-green range of spectrum. On the other hand, because the spectral absorption of the 380 nm photoreceptor and the spectral sensitivity of the 510 nm photoreceptor overlapped, light absorption by the UV sensitizing pigment and the bluegreen sensitive rhodopsin of the 510 nm photoreceptor is hindered by the 380 nm photoreceptor. Consequently, UV sensitivity is diminished as a small sensitive peak, and the blue-green sensitivity is also altered, so that the spectral sensitivity curve ts best with the 510 nm theoretical visual pigment absorption curve.

ARTICLE IN PRESS
W.-Y. Wu et al. / Journal of Insect Physiology 53 (2007) 509516 515

4.2. Color preference of the oriental fruit y Previous studies on another tephritid pest, the apple maggot y (Rhagoletis pomonella), demonstrated that the intensity contrast of dark host against bright background plays a more important role in the ys host detection than host colour (Agee, 1985; Owens and Prokopy, 1986). Contrarily, a bright host (e.g. colored with yellow or white) is more attractive to the oriental fruit y when tested with different-color spheres (Vargas et al., 1991), suggesting that different strategies are used for seeking host targets between the two fruit y species. When comparing the reectance spectra of the colored papers and their attractiveness for the oriental fruit y (Figs. 1 and 8), it is evident that several spectral cues are involved with the attractiveness. The green colored paper that showed the higher attractiveness among the colored papers has a broader and higher reectance, indicating that chromatic stimulation of the spectra within 500570 nm contains crucial cues for attracting the oriental fruit y. This is also supported by the fact that orange colored paper with a reectance spectrum slightly higher than that of red colored paper, in the range of 470630 nm, has a signicantly stronger attractiveness to the ies than the red colored paper. In addition, the white colored paper used in our experiment has the broadest reectance spectrum but the lowest attractiveness to the ies, even less than the black colored paper, and the green colored paper has a similar attractive effect as the yellow colored paper, which has a much higher reectance within 500570 nm than the green colored paper. Thus, it is reasonable to hypothesize that there is a negative cue in the spectrum to reduce the attractiveness. The UV-sensitive peaks seen in the recorded spectral sensitivity curves of the y initially suggested that negative cues might be present in this spectral region. However, the preference test indicated that UV reection enhanced attractiveness, as did the green reection (500570 nm). The UV-positive effect could be attributed to the UV sensitivity of the dual-peak photoreceptors as well as the 370 nm and the 380 nm photoreceptors. A nal possibility for the location of the negative cue could be within the visual sensitive range of 380500 nm, for which no corresponding photoreceptor was recorded in this study. This negative effect could be caused by opponent mechanisms, or related to an undiscovered blue photoreceptor (sensitivity range between 400 and 500 nm), as suggested by a previous ERG study on the oriental fruit y (Wu, 1989) and the evidence shown in other dipteran ies (Hardie, 1986). The green-positive effect could be attributed to the green sensitivity of the 510 nm photoreceptor, or the related opponent mechanisms as mentioned above. Since the action spectra of the photoreceptors overlapped each other, genetic mutants such as photoreceptor-defective strains or RNAi knock down of individual photopigments would be required to conrm which photoreceptors were actually involved with color preference. Nevertheless, the

chromatic cues of the color preference obtained by applying the fundamental physiological information are helpful in studying the visual ecology of the oriental fruit y, and provide a useful tool for developing pest control methods. While various yellow traps have been widely used in eld trapping, most of the yellow colors were determined by human perception without precisely quantifying the action spectrum of the pest and concerning the negative effect. Our results suggest that a trap with sufcient reection within 300380 nm and 500570 nm but without reection within 380500 nm will be more effective in attracting the oriental fruit y. Acknowledgment We thank the anonymous reviewers for insightful corrections and improvements. We are also most grateful to Dr. Kuang-Hui Lu and his lab for their generous y stocks. This work was supported by a grant (94 AS-13.2.1BQ-BG) from the BAPHIQ and a grant (NSC 91-2313-B005-117) from National Science Council, Taiwan. References
Agee, H.R., 1985. Spectral response of the compound eye of the wild and laboratory-reared apple maggot y, Rhagoletis pomonella. Journal of Agricultural Entomology 2, 147154. Alyokhin, A.V., Messing, R.H., Duan, J.J., 2000. Visual and olfactory stimuli and fruit maturity affect trap captures of oriental fruit ies (Diptera: Tephritidae). Journal of Economic Entomology 93, 644649. Chen, C.C., Dong, Y.J., 2001. Evaluation of trapping effectiveness of the improved McPhail trap for oriental fruit y (Bactrocera dorsalis) (Diptera: Tephritidae). Formosan Entomologist 21, 6575 (in Chinese). Christenson, L.C., Foote, R.H., 1960. Biology of fruit ies. Annual Review of Entomology 5, 171192. Cornelius, M.L., Duan, J.J., Messing, R.H., 1999. Visual stimuli and the response of female oriental fruit ies (Diptera: Tephritidae) to fruitmimicking traps. Journal of Economic Entomology 92, 121129. Corneluis, M.L., Duan, J.J., Messing, R.H., 2000a. Volatile host fruit odors as attractants for the oriental fruit y (Diptera: Tephritidae). Journal of Economic Entomology 93, 93100. Corneluis, M.L., Neregel, L., Duan, J.J., Messing, R.H., 2000b. Responses of female oriental fruit ies (Diptera: Tephritidae) to protein and host fruit odors in eld cage and open eld tests. Environmental Entomology 29, 1419. Ebrey, T.G., Honig, B., 1977. New wavelength dependent visual pigment nomograms. Vision Research 17, 147151. Haramoto, F.H., Bess, H.A., 1970. Recent studies on the abundance of the oriental and Mediterreanean fruit ies and the status of their parasites. Proceeding of the Hawaiian Entomological Society 20, 551566. Hardie, R.C., 1979. Electrophysiological analysis of y retina. I: comparative properties of R1-6 and R7 and 8. Journal of Comparative Physiology A 129, 1933. Hardie, R.C., 1986. The photoreceptor array of the dipteran retina. Trends in Neurosciences 9, 419423. Hardie, R.C., Franceschini, N., Mclntyre, P.D., 1979. Electrophysiological analysis of y retina II. Spectral and polarisation sensitivity in R7 and R8. Journal of Comparative Physiology A 133, 2339. Heath, R.R., Epsky, N.D., Landolt, P.J., Sivinski, J., 1993. Development of attractants for monitoring Caribbean fruit ies (Diptera: Tephritidae). Florida Entomologist 76, 233244.

ARTICLE IN PRESS
516 W.-Y. Wu et al. / Journal of Insect Physiology 53 (2007) 509516 Owens, E.D., Prokopy, R.J., 1986. Relationship between reectance spectra of host plant surfaces and visual detection of host fruit by Rhagoletis pomonella ies. Physiological Entomology 11, 297307. Prokopy, R.J., Owens, E.D., 1983. Visual detection of plants by herbivorous insects. Annual Review of Entomology 28, 337364. Prokey, R.J., Green, T.A., Vargas, R.I., 1990. Dacus dorsalis ies can learn to nd and accept host fruit. Journal of Insect Behavior 3, 663672. Raptopoulos, D., Haniotakis, G., Koutsaftikis, A., Kelly, D., Mavraganis, V., 1995. Biological activity of chemicals identied from extracts and volatiles of male Rhagoletis cerasi. Journal of Chemical Ecology 21, 12871297. Roessler, Y., 1989. Insecticidal bait and cover sprays. In: Robinson, A.S., Hooper, G. (Eds.), World Crop Pests, Vol. 3B, Fruit Flies, their Biology, Natural Enemies and Control. Elsevier, Amsterdam, pp. 329336. Sivinski, J., 1990. Colored spherical traps for capture of Caribbean fruit y, Anastrepha suspensa. Florida Entomologist 73, 123128. Vargas, R.I., Stark, J.D., Prokopy, R.J., 1991. Response of oriental fruit y (Diptera: Tephritidae) and associated parasitoids (Hymenoptera: Braconidae) to different-color spheres. Journal of Economic Entomology 84, 15031507. Wu, C.Y., 1989. Receptors in insect II. Electroretinogram of the compound eye in the oriental fruit y (Dacus dorsalis Hendel). Bulletin of the Institute of Zoology, Academia Sinica 28, 713. Wu, C.Y., Chang, C.S., Tung, L.C., Lin, J.T., 1985. Receptors in insect I. The ne structure of the compound eye of the oriental fruit y Dacus dorsalis Hendel. Bulletin of the Institute of Zoology, Academia Sinica 24, 2738. Horridge, G.A., Mimura, K., 1975. Fly photoreceptors I. Physical separation of two visual pigments in Calliphora retinula cells 16. Proceedings of the Royal Society of London B 190, 211224. Horridge, G.A., Mimura, K., Tsukahara, Y., 1975. Fly photoreceptors II. Spectral and polarized light sensitivity in the drone y Eristalis. Proceedings of the Royal Society of London B 190, 225237. Jang, E.B., Light, D.M., 1991. Behavioral responses of female oriental fruit ies to the odor of papayas at three ripeness stages in a laboratory ight tunnel (Diptera: Tephritidae). Journal of Insect Behavior 4, 751762. Katsoyannos, B.I., 1989. Response to shape, size and color. In: Robinson, A.S., Hooper, G. (Eds.), World Crop Pests, Vol. 3A, Fruit Flies, their Biology, Natural Enemies and Control. Elsevier, Amsterdam, pp. 307324. Kirschfeld, K., Franceschini, N., Minke, B., 1977. Evidence for a sensitizing pigment in y photoreceptors. Nature 269, 386390. Light, D.M., Jang, E.B., Flath, R.A., 1992. Electroantennogram responses of the Mediterranean fruit y, Ceratitis capitata, to the volatile constituents of nectarines. Entomologia Experimentalis et Applicata 63, 1326. Liu, Y.C., 1981. A review on studies of the oriental fruit y, Dacus dorsalis Hendel in Taiwan. Bulletin of the Society of Entomology 16, 926 (in Chinese). Loew, E.R., 1975. CO2-induced changes in the ERG of the y Sarcophaga bullata. A component analysis. Journal of Insect Physiology 21, 181197. Manzel, R., 1975. Colour receptors in insects. In: Horridge, G.A. (Ed.), The Compound Eye and Vision of Insects. Oxford University Press, London, pp. 121153.