- Email: [email protected]
Patterns of chromatic information processing in the lobula of the honeybee, Apis mellifera L.
ARTICLE IN PRESS
Journal of Insect Physiology 50 (2004) 913–925 www.elsevier.com/locate/ip
Patterns of chromatic information processing in the lobula of the honeybee, Apis mellifera L. En-Cheng Yang, Hsiao-Chun Lin, Yu-Shan Hung a
Department of Entomology, National Chung Hsing University, 250 Kuo Kuang Rd., Taichung 402, Taiwan Received 14 March 2004; received in revised form 24 June 2004; accepted 28 June 2004
Abstract The honeybee, Apis mellifera L., is one of the living creatures that has its colour vision proven through behavioural tests. Previous studies of honeybee colour vision has emphasized the relationship between the spectral sensitivities of photoreceptors and colour discrimination behaviour. The current understanding of the neural mechanisms of bee colour vision is, however, rather limited. The present study surveyed the patterns of chromatic information processing of visual neurons in the lobula of the honeybee, using intracellular recording stimulated by three light-emitting diodes, whose emission spectra approximately match the spectral sensitivity peaks of the honeybee. The recorded visual neurons can be divided into two groups: non-colour opponent cells and colour opponent cells. The non-colour opponent cells comprise six types of broad-band neurons and four response types of narrow-band neurons. The former might detect brightness of the environment or function as chromatic input channels, and the latter might supply specific chromatic input. Amongst the colour opponent cells, the principal neural mechanism of colour vision, eight response types were recorded. The receptive fields of these neurons were not centre surround as observed in primates. Some recorded neurons with tonic post-stimulus responses were observed, however, suggesting temporal defined spectral opponency may be part of the colour-coding mechanisms. r 2004 Elsevier Ltd. All rights reserved. Keywords: Colour vision; Colour coding; Colour opponency; Honeybee; Apis mellifera
1. Introduction Colour is a meaningful signal for many animals, both vertebrate and invertebrate. Because colour discrimination is not present in all aspects of visual orientation (Lehrer, 1987; Zhang and Srinivasan, 1993), and because of the difficulty of behavioural training, so far, the only a few invertebrate species, e.g. bees (Frisch, 1914) and wasps (Chittka et al., 1992), flies (Troje, 1993; Fukushi, 1994), butterfly (for example, Kolb and Scherer, 1982; Kelber and Pfaff, 1999; Kinoshita et al., 1999) and stomatopods (Marshall et al., 1996), have been studied sufficiently carefully to draw the conclusion that they have true colour vision (rev. Autrum and Corresponding author. Tel./fax: +886-4-22856079
E-mail address: [email protected] (E.-C. Yang). 0022-1910/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.jinsphys.2004.06.010
Thomas, 1973; Goldsmith and Bernard, 1974; Menzel, 1979; Barth, 1985; Menzel and Backhaus, 1989; Kelber et al., 2003). Thanks to the successes of intracellular recording and behavioural reinforcement training, it has been convincingly demonstrated that the honeybee possesses a trichromatic visual system with UV (335 nm), blue (435 nm) and green (540 nm) receptors (Autrum and von Zwehl, 1964; Menzel and Blakers, 1976; Menzel et al., 1986; Peitsch et al., 1992). Behavioural investigations of colour vision in the honeybee have been carried out intensively since Karl von Frisch first demonstrated that bees see colour objects as a visual quality different from brightness, at least in the behavioural contexts of feeding and homing (von Frisch, 1914; rev. von Frisch, 1967). Behavioural experiments also demonstrated that, like humans, bees are able to correctly identify the colours of objects in
ARTICLE IN PRESS 914
E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
spite of dramatic changes in the spectral content of the illumination, that is to say they exhibit colour constancy (Neumeyer, 1981; Werner et al., 1988). It has also been demonstrated that the hue perceived by bees when viewing a given colour stimulus is altered after adaptation to a different colour stimulus. This phenomenon is known as successive colour contrast (Neumeyer, 1981; Dyer and Chittka, 2004). Colour perception is also influenced by the presence of other colours in the surround, an effect described as simultaneous colour contrast by Neumeyer (1980). Evidence from behavioural experiments has demonstrated that some behaviours related to particular visual functions is driven by a particular wavelength-specific neural pathway due to one of the three types of receptors. For example, honeybees use their UV receptors to detect polarized light and to determine orientation for navigation (Menzel and Snyder, 1974), and they use their green receptors for optomotor response (Kaiser, 1974, 1975) and for flight distance estimation (Chittka and Tautz, 2003). Studies on the form vision of the honeybee also show that the green receptors are the dominant signal source in the orientation-sensitive neurons (Srinivasan, 1994; Giger and Srinivasan, 1996). In contrast to the abundant information on behavioural experiments, physiological studies of honeybee colour vision are relatively few. Kien and Menzel (1977a, b) first reported three classes of colour-coding neurons in the optic lobes (mainly in the medulla and a few in the lobula) of the honeybee: broad-band neurons, narrow-band neurons and colour opponent neurons. The broad-band neurons can respond to light from across the broader spectrum, suggesting that they receive inputs from two or three receptor types. The narrow-band neurons are only sensitive to wavelengths within the spectral sensitivity of one photoreceptor type. The colour opponent neurons are excited by some wavelengths but are inhibited by others. From earlier studies in vertebrate visual systems we know that colour opponency is one of the most important neural mechanisms for processing colour information (De Valois, 1973). Previous studies of the optic lobes in the honeybee also showed similar colour opponent response patterns, however, the receptive fields of the recorded colour opponent neurons were not spatially antagonistic, i.e. did not have a centre-surround configuration, as some of those of vertebrates (Kien and Menzel, 1977a, b). In addition, by means of intracellular recording and staining, Hertel (1980) reported the response tunings of colour-coding neurons in the medulla and lobula. By using a heterochromatic flicker test for the colour opponent neurons of the lobula, Riehle (1981) provided neurophysiological evidence for successive colour contrast in bees.
Neural mechanisms have been an historic focus in primate colour vision systems. In invertebrates, most of the early studies of colour vision emphasized colour discrimination behaviour and the interaction between different spectral types of photoreceptors, while nowadays the trend is to focus on higher order neural mechanisms (Kelber et al., 2003). To study the neural mechanisms of the colour vision of a honeybee, it is important to see how the spectral signals are processed in the visual neurons. Only by identifying the response patterns of the light-sensitive neurons will we then be able to further investigate the roles each type of neuron plays in colour processing. In the present study we surveyed the spectral response patterns of the honeybee’s visual neurons with conventional intracellular recordings in the lobula. Together with previous studies, our results show that in total 19 possible combinations of spectral signals from the three receptors can be recorded from the light-sensitive neurons in the lobula.
2. Materials and Methods 2.1. Animals Experiments were carried out on adult honeybee workers (Apis mellifera L.) which were reared on the campus of the National Chung Hsing University in Taiwan. These honeybees were caught at the hive entrance while they were just returning to the hive. All honeybees were used immediately after capture. The honeybees were kept at room temperature (25721C) to sustain their physiological condition. 2.2. Preparation The honeybees were anaesthetized by placing them in a refrigerator at 41C for about 5 min. The thorax was then mounted on a plastic plate, which was fixed to a brass pedestal, with a beeswax/resin (3:1) mixture, so that the head and thorax were rigidly secured. The abdomen was free to perform ventilatory movements. The legs were amputated, and the wings were waxed together. The vertex of the head capsule between the left compound eye and the ocellus was removed to expose the optic lobe to admit recording electrodes. 2.3. Electrophysiological recording Electrophysiological recording was performed in a dark room to keep the honeybees dark-adapted. Microelectrodes, made of aluminosilicate glass capillaries (AF100-68-10, Sutter Instruments) and pulled on a micropipette puller (Flaming/Brown P-97, Shutter Instruments), had a resistance of between 70 and 90 MO when filled with 2 M potassium acetate and measured in
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
the brain tissue. The microelectrode was lowered vertically by a micromanipulator (MWS-32, Narishige) to penetrate the lobula. The manipulation was observed via a stereomicroscope (GZ6, Leica) to ensure that the microelectrode was accurately inserted into the lobula. The brass pedestal, with the honeybee mounted, and the micromanipulator were secured onto a vibration isolating workstation (LW3030-OPT, Newport) by means of magnetic bases. The indifferent electrode was a silver wire inserted into the thorax. The electrophysiological signal was amplified 10 (Neuroprobe, Model 1600, AM Systems). The signal was then displayed on an oscilloscope (GOS-620FG, GW-Instek), simultaneously acquired by a data-acquisition system (PCI-6024E, National Instruments) and stored in a personal computer for further processing. The control program for the acquisition system and the program for analysing the recorded data were both written in LabView (National Instruments). The recorded waveforms were digitized at sampling rate of 5 kHz. 2.4. Visual stimulus A conventional flash method was used to stimulate the recorded visual neurons. Three light-emitting diodes (LEDs) were used as the light source. The emission spectra of the three LEDs are shown in Fig. 1, and their peak wavelengths are: 370 nm (UV), 435 nm (blue) and 560 nm (green). These peak wavelengths were chosen according to match as closely as possible the spectral sensitivity of three types of honeybee photoreceptors (Peitsch et al., 1992). A multi-channel controller provided suitable voltages to drive each of the three LEDs. The intensities of the peak wavelengths of the UV, blue and green LEDs were 4.44 1013, 8.58 1013,
Fig. 1. Relative spectral radiant power distribution of the three LEDs used as visual stimulus in the experiment. The wavelengths of peak emission are at 370, 435 and 560 nm.
915
and 2.10 1013 photons/cm2/s, respectively. The light source was mounted on a Cardan arm perimeter, which could be accurately positioned at the point of maximum sensitivity of the recorded cell. Because only one of the three LEDs could be mounted on the Cardan arm, the LED was replaced when a colour stimulation change was needed. The angles subtended at the eye of the three light sources were: 3.31 for UV and 3.51 for both blue and green. The stimulating flash lasted for 1 s, and the interval between flashes was 10 s. Both the duration and the interval were controlled with an isolated pulse stimulator (Model 2100, A-M Systems). 2.5. Calibration A miniature fibre optic spectrometer (S2000, Ocean Optics) was used to determine the emission spectra of the LEDs. A radiometer (S370, UDT Instruments) with detectors (Model 222, 260, UDT Instruments) was employed to measure the light flux at the position of the eye. The energy measured at different monochromatic wavelengths was transformed into photons/cm2/s.
3. Results 3.1. General response properties to flash The spontaneous activity frequency and the amplitude of the recorded action potentials was within the range 0–80 Hz and 5–60 mV, respectively. Only those cells with spike amplitudes of more than 10 mV and basalline shifting of less than 10 mV were analysed in this report. More than 150 cells were recorded from the lobula. Among them, 97 cells were sensitive to flash stimuli. The receptive fields of the 97 cells were large (X601), and no centre-surround configurations were found. Kien and Menzel (1977a) divided the flashresponse profiles of the high-order visual neuron into three stages: on-effect, sustained response and off-effect. Both on- and off-effects are phasic, and the sustained response is tonic. Most of the 97 cells showed excitatory (67 cells) or inhibitory (13 cells) phasic responses, and only 17 cells responded tonically. The phasic response could be related to enhancing colour contrast, while the tonic response is presumed to be responsible for the sustained property of colour stimuli (Hertel and Maronde, 1987). Therefore, the analysis of neuronal responses to flashes is focused on the tonic response, and consequently the 34 cells showing only a phasic response are not considered in this report. Based on the tonic responses to different colour flash stimuli, the recorded neurons can be divided into two groups: non-colour opponent cells (45 cells) and colour opponent cells (10 cells).
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
916
3.2. Non-colour opponent cells We used the nomenclature of Kien and Menzel (1977a, b) to describe the 45 non-colour opponent cells. Thus, the 33 that cells were sensitive to more than one colour flash stimulus are referred to here as broad-band neurons. The remaining 12 cells were sensitive to one colour flash stimulus only and so are referred to as narrow-band neurons. Based on their response patterns, the broad-band neurons can be further categorized as:
Table 1 Non-colour opponent neural types, response patterns and numbers of colour-coding neurons recorded from the lobula of the honeybee Neural types Broad band neurons Flash-sensitive All excitatory All inhibitory Dual excitatory Dual inhibitory Narrow-band neurons Single excitatory Single inhibitory
Response patterns
No. of recording
Only phasic response UV+/B+/G+ UV /B /G B+/G+ UV /B B /G
34 10 8 5 8 2
UV+ B+ UV B
3 3 2 4
all excitatory neurons, all inhibitory neuron dual excitatory neuron, and dual inhibitory neuron (Table 1). Among the 33 broad-band neurons, 18 cells were recorded as sensitive to all three colour flash stimuli, including the ‘all excitatory’ neuron (UV+/B+/ G+, Fig. 2a) and the ‘all inhibitory neuron’ (UV /B / G , Fig. 2b). These two types of response patterns have been previously reported (Kien and Menzel, 1977a). The other 15 cells respond to two of the three colours: one type is the dual excitatory neurons (B+/G+, Fig. 3a) and these are also two types of inhibitory neurons (UV /B , Fig. 3b; B /G , Fig. 3c). These types of neurons have not been previously reported. Neither the response pattern of UV+/G+nor UV /G was recorded in our study. Kien and Menzel (1977a) once reported a neuron with a response pattern of UV+/ G+, and claimed that a neuron receiving inputs with the same sign from two non-overlapping receptors could in no way discriminate colour. It could, however, feed into a spectrally antagonistic neuron afterwards. The narrow-band neurons are categorized as: single excitatory neuron and single inhibitory neuron. Both UV-sensitive (UV+, Fig. 4a; UV , Fig. 4c) and blue-sensitive (B+, Fig. 4b; B , Fig. 4d) neurons were recorded; however, no green-sensitive neurons (neither G+ nor G ) was found. Green sensitive narrow-band neurons were previously reported (G+, Kien and Menzel, 1977a), but other types of
Fig. 2. The response patterns of flash-sensitive neurons elicited with three spectral light stimuli (UV, blue and green). (a) A neuron with phasic excitatory responses while stimulated with light on and off. (b) A different neuron showing phasic responses to light on and off. The horizontal bar under each response depicts the duration of a 1 s flash stimulus; the vertical bar at the right-hand side depicts response amplitude of 10 mV. Response histograms to 10 stimulations are shown below the response waveforms.
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
917
Fig. 3. The response patterns of broad-band neurons elicited with three spectral light stimuli (UV, blue and green). Dual excitatory neuron: (a) B+/ G+; dual inhibitory neurons: (b) UV /B , (c) B /G . The scales of the horizontal and vertical bars are the same as the depictions in Fig. 2.
narrow-band neurons (UV-and blue-sensitive only) are reported here for the first time. Numbers of recorded cells for each type of non-colour opponent neurons are listed in Table 1. 3.3. Colour opponent cells A total of 10 recorded cells with seven types of response patterns are colour opponent cells. Based on the opponent patterns these cells can be further categorized as: single opponent neuron, dual excitatory opponent neuron and dual inhibitory opponent neuron (Table 2). For the single opponent neurons, only the neurons inhibited by short wavelengths but excited by long wavelengths were recorded: UV /B+ (Fig. 5a), UV /G+ (Fig. 5b) and B /G+ (Fig. 5c); all the counter parts of the three opponent types, UV+/B , UV+/G and B+/G , were not found. Among the
three types of single opponent cells found, only the UV /G+ was previously reported (Hertel and Maronde, 1987). For the dual opponent neurons, three types of dual excitatory opponent neurons, UV+/B+/G (Fig. 6a), UV+/B /G+ (Fig. 6b) and UV /B+/G+ (Fig. 6c), and one type of dual inhibitory neuron, UV /B+/G (Fig. 7), was recorded. Previous studies on bee colour opponent cells indicated that UV+/B / G+ and UV+/B /G (and their counter parts UV /B+/G and UV /B+/G+) are the two main classes of colour opponent cells in the bee brain (Kien and Menzel, 1977b; Hertel, 1980; Menzel and Backhaus, 1989). Together with our present study only the counterpart of UV+/B+/G (i.e. UV /B /G+) was not found so far. Numbers of recorded cells for each type of colour opponent neurons are listed in Table 2.
ARTICLE IN PRESS 918
E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
Fig. 4. The response patterns of narrow-band neurons elicited with three spectral light stimuli (UV, blue and green). UV-sensitive neurons: (a) UV+, (c) UV ; blue sensitive neurons: (b) B+, (d) B . The scales of the horizontal and vertical bars are the same as the depictions in Fig. 2.
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925 Table 2 Colour opponent neural types, response patterns and numbers of colour-coding neurons recorded from the lobula of the honeybee Neural types Colour opponent neurons Single opponent
Dual excitatory opponent
Dual inhibitory opponent
Response patterns
No. of recording
UV /B+ UV /G+ B /G+ UV+/B+/G UV+/B /G+ UV /B+/G+ UV /B+/G
2 1 1 1 1 3 1
919
3.4. Post-stimulus response patterns During our recording some cells responded following the end of the colour flash. Two of six broad-band neurons are shown as examples in Fig. 8. The poststimulus responses were either excitatory (Fig. 8a) or inhibitory (Fig. 8b). The post-stimulus response patterns were not only found in non-colour opponent cells but also in colour opponent cells. Two colour opponent cells, as per examples shown in Figs. 6b and 7, exhibit strong post-stimulus responses to the three colour stimuli. Some colour opponent cells showed strong excitatory post-stimulus responses to the UV, blue and
Fig. 5. The response patterns of single opponent neurons elicited with three spectral light stimuli (UV, blue and green): (a) UV /B+; (b) UV /G+ and (c) B /G+. The scales of the horizontal and vertical bars are the same as the depictions in Fig. 2.
ARTICLE IN PRESS 920
E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
Fig. 6. The response patterns of three types of dual excitatory opponent neurons: (a) UV+/B+/G , (b) UV+/B /G+, and (c) UV /B+/G+. The scales of the horizontal and vertical bars are the same as the depictions in Fig. 2. Note that (b) shows strong post-stimulus responses to the three spectral stimuli: UV+/B+/G+.
Fig. 7. The response pattern of one type of dual inhibitory opponent neurons: UV /B+/G . The scales of the horizontal and vertical bars are the same as the depictions in Fig. 2. Note that the responses of this neuron shows strong post-stimulus responses to the three spectral stimuli: UV+/B /G+.
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
921
Fig. 8. The response patterns of two broad-band neurons with post-stimulus response elicited with three spectral light stimuli (UV, blue and green). (a) Tonic flash responses (UV+/G+), excitatory post-stimulus responses (UV+/B+/G+); (b) tonic flash responses (UV /B /G ), inhibitory post-stimulus responses (UV+/B /G+). The scales of the horizontal and vertical bars are the same as the depictions in Fig. 2.
green lights (Fig. 6b), and some even showed antagonistical responses. For example, the post-stimulus response pattern of the colour opponent cell shown in Fig. 7 was excited by UV and green lights but inhibited by blue, making the colour opponency of the poststimulus response UV+/B /G+, which is opposite to its tonic response patterns (UV /B+/G ).
4. Discussion 4.1. Response patterns of colour-coding neurons In the honeybee lobula, visual neurons can be divided into two groups according to whether or not they generate spontaneous responses (Hertel, 1980). Erber and Menzel (1977) suggested that the responses of highorder visual neurons elicited by different colour stimuli were similar, except for the frequency of spontaneous responses. In this present study all the 97 recorded neurons generated spontaneous responses, most of them (69%) with an on- or off-transient response when stimulated by flashed light. This result is similar to a previous study by Hertel (1980), in which a total of 88 visual neurons were recorded, and 52% of them gave
on-transient responses. In contrast to the transient response, the colour-coding neurons, which are of the most interest to us, should be able to sustain their responses to a prolonged colour stimulus. About 57% (55 neurons) in our recorded neurons had a tonic response to colour flashes, some of them providing both transient and tonic responses when stimulated by a flash. Previous recordings by Hertel (1980) showed that only 16% of recorded neurons had tonic responses. Since our recordings were confined to the lobula, and most recording sites of the previous study by Hertel (1980) were in the medulla, and only few cells were recorded from the lobula, this might explain the difference in the tonic response patterns of colour-coding neurons between the two studies. Table 3 shows the comparison of colour-coding neurons between previous studies and our results. Sixteen response patterns were recorded, of which seven of them are new discoveries. Together with the 12 response patterns found by previous studies (Kien and Menzel, 1977a, b; Hertel and Maronde, 1987), a total of 19 response patterns, distributed among all the categorized neural types (Tables 1 and 2), were observed. Although seven response patterns were still not revealed,
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
922
Table 3 The response patterns of recorded colour-coding neurons in the optic lobes of the honeybee, reported from this present study and from previous studies Response patterns Broad band UV+/B+/G+ UV /B /G UV+/B+a UV+/G+ B+/G+ UV /B UV /G a B /G Narrow band UV+ B+ G+ UV B G a Colour opponent UV+/B a UV /B+ UV+/G a UV /G+ B+/G a B /G+ UV+/B+/G UV+/B /G+ UV+/B /G UV /B /G+a UV /B+/G UV /B+/G+
Recorded neurons
Previous studies
— —
—(a) —(a) —(a)
— — — — — — —
—(b) —(b) —(b) —(b)
— — — — —
— —
—(c)
—(b) —(b) —(b) —(b)
a The response patterns have not been found yet and are only hypothesized. (a) Kien and Menzel (1977a); (b) Kien and Menzel (1977b) and (c) Hertel and Maronde (1987).
a close examination of these missing patterns (e.g. UV / B /G+) shows that their counter parts (e.g. UV+/B+/ G ) all were recorded (Table 3). 4.2. Neural mechanisms for honeybee colour vision Menzel (1979, 1985) proposed roles that each type of response pattern may play in processing the colour signals of insect vision. He suggested that the broadband neurons are designated as luminosity-type neurons, which are not involved in colour vision but are sensitive to the light intensity of the environment, and that the narrow-band neurons and colour opponent cells are related to colour vision and extract colour signals with different response patterns. In primates, the analogues of the all excitatory (UV+/B+/G+) and all inhibitory (UV /B /G ) neurons were suggested as to process signals for achromatic sensation (Zrenner et al., 1990). Kien and Menzel (1977a) presented three
types of broad-band neurons, in addition to UV+/B+/ G+ and UV /B /G , the response pattern of UV+/ G+ providing a type of neuron that is unable to distinguish UV from green. Our results show that the neurons recorded with dual inputs were from UV/blue or blue/green receptors. These dual-input broad-band neurons may not only be related to sensing luminosity but may, together with other colour-coding neurons, form input channels to higher-order neurons showing colour opponency. Narrow-band neurons are sensitive to the light stimulus within a restricted range of the spectrum, and their spectral sensitivities are usually similar to some photoreceptor type (Menzel, 1979). However, the spectral sensitivities of some narrow-band neurons recorded by Kien and Menzel (1977b) were narrower when compared to those of the photoreceptors. Although we did not measure the spectral sensitivity of our recorded neurons, our results show that only UVand blue-sensitive neurons were recorded, suggesting these neurons received inputs from UV and blue receptors only. The reason for no green-sensitive neuron being recorded could be due to the difference in recording sites. Our recording microelectrode was inserted vertically from the dorsal part of lobula, while Kien and Menzel (1977b) mainly recorded from the ventral part of the optic lobes. Because there are three main neural tracts for projections from the optic lobes to the protocerebrum in the visual system of a honeybee (Ehmer and Gronenberg, 2002), it is possible that the UV and blue narrow-band neurons might be in a tract different from the green-sensitive neurons. Neuroanatomical study of the parallel visual pathways in Diptera indicates that the retinotopic columns in the lamina comprise two pathways: the spectrally insensitive channel (R1-6/L1, 2, green sensitive) and the trichromatic or tetrachromatic channel (R7, 8, L3). The two channels separately supply two subsets of retinopotic neurons throughout the entire visual system according to their relationships with retina/lamina afferents (Strausfeld and Lee, 1991). To verify the separated channels in the visual system of the honeybee, further recordings with dye injection to reveal the neural profiles are necessary. It must be noted that the colour opponent cells shown in Kien and Menzel (1977b) were recorded from the medulla, rather than the lobula. Thus they suggested that the medulla was the region of the optic lobe in which colour signals were processed. Based on the two colour opponent types only, two counterparts for each type (type A: UV /B+/G+ and UV+/B /G ; type B: UV /B+/G and UV+/B /G+), Backhaus (1991) was able to construct a colour opponent coding model to explain the results obtained from the colour discrimination behaviour of the honeybee (von Helversen, 1972). Our results show that the colour
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
opponent cells are various, including the single opponent cells and the dual opponent cells (Table 2). Since the counterparts of all the missing types were recorded, it is very likely that all the 12 assumed types of colour opponent cells can be recorded from the lobula, although one certainly does not need all the types and further recordings are necessary to verify this point. One might wonder why so many different types of colour opponent cells exist in honeybees when only two are necessary to explain colour discrimination. It is reasonable to hypothesize that the single opponent cells may form the inputs for the three types of dual opponent cells, including a third pair (i.e. type C: UV+/B+/G and UV /B /G+). In addition, some possible functions related to particular visual channels have been identified by behavioural experiments, and our recorded cell types might be involved. For instance, a UV /G+ neuron and its counterpart might be part of a visual channel that provides UV–green contrast in order to discriminate an object against the background of a bright sky; such a channel might be employed when a bee uses landmarks to orient (Mo¨ller, 2002). 4.3. Colour opponent cells of primates and honeybees On the photoreceptor level, both primates and honeybees have three spectral types of photoreceptors, and thus both systems for colour vision are based on the theory of trichromacy. On the neural level, both primates and honeybees have colour opponent coding neurons, and therefore they can transfer the perceived spectral signals to colour signals. With colour vision, primates and bees are able to perform colour constancy and simultaneous colour contrast in behavioural experiments. Although trichromatic systems offer the animals to be able to discriminate colours, at the level of neural mechanisms for colour coding, the receptive fields of colour opponent cells of honeybees are different from those of primates. In contrast to the centre-surround receptive fields observed from primate visual neurons, to date no such receptive field is found from any recorded colour opponent cell of honeybees. Without the centresurround configuration, the colour opponent cells of honeybees would not be able to compare the contrast of neighbouring colours at the same time. Menzel and Backhaus (1989) proposed that simultaneous colour contrast might only exist for particular patterns of colour arrangements. Other colour contrast phenomena may be based solely on successive contrast effects, i.e. based on temporary sampling of contrast values. To determine the colour of an object independently of the chromatic illumination, the nervous system of honeybee must compare the wavelength composition of light reflected by an object with respect to the wavelength composition reflected from the surrounding
923
area. The neurons with post-stimulus response patterns reported in this present study are probably related to this hypothesis, and may provide the neural basis for the honeybee to compute the colour contrast when scanning over areas (Menzel and Backhaus, 1989). For the variety of response patterns, in contrast to only six patterns being found in the primates visual system: R+/G (R /G+), G+/R (G /R+) and B+/R G (B /R+/G+), and their counter parts (Zrenner et al., 1990), there are at least eight patterns in the honeybee’s visual system. This difference could be attributed to the evolutionary background of the two trichromatic systems. Whereas the rods and the shortwavelength (S) cones are phylogenetically older, the long (L) and middle-wavelength (M) cones evolved from a common ancestoral pigment only about 35 million years ago (Nathans, 1989; Osorio and Vorobyev, 1996; Bowmaker, 1998; Gegenfurtner, 2003). It has been known that primates have more red and green opponent cells than others (Zrenner et al., 1990). In contrast to the primates, the basic UV–blue–green trichromacy appears to date back to the Devonian ancestor of all pterygote insects, and each of the three visual pigments diverged very early in evolution (Briscoe and Chittka, 2001). The interaction of chromatic information from different types of photoreceptor in the insect visual system might have been developled since then. One might draw the inference that colour coding on the neuronal level might be more complex in bees than in humans (or primates). Previous behavioural work in bees by Backhaus and Menzel (Menzel and Backhaus, 1989; Backhaus, 1991) derived two axes for a psychophysical colour space which correspond to the colour opponent neurons reported by Kien and Menzel (1977b). While behavioural work seems to show that there is two dimensionality of colour space, colour coding on the neuronal level seems to be based on even more mechanisms than in humans.
Acknowledgements We would like to thank L. Chittka, D. Osorio and T. Maddess for reading and correcting our manuscript. Comments from two anonymous referees greatly improved our manuscript. This work was supported by a project grant (NSC 89-2313-13-005-166) from National Science Council, Taiwan, ROC.
References Autrum, H., Thomas, I., 1973. Comparative physiology of colour vision in animals. In: Jung, R. (Ed.), Handbook of Sensory Physiology, vol. 7/3A. Springer, Berlin, pp. 661–692.
ARTICLE IN PRESS 924
E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
Autrum, H., von Zwehl, V., 1964. Die Spektrale Empfindlichkeit Einzelner Sehellen des Bienenauges. Zeitschrift fu¨r vergleichende Physiologie 48, 357–384. Backhaus, W., 1991. Colour opponent coding in the visual system of the honeybee. Vision Research 31, 1381–1397. Barth, F.G., 1985. Insect and Flowers—The Biology of a partnership (Translated by M. A. Biederman-Torson). Princeton University Press, Englewood Cliffs, NJ, p. 122. Bowmaker, J.K., 1998. Evolution of colour vision in vertebrates. Eye 12, 541–547. Briscoe, A.D., Chittka, L., 2001. The evolution of colour vision in insects. Annual Review of Entomology 46, 471–510. Chittka, L., Tautz, J., 2003. The spectral input to honeybee visual odometry. Journal of Experimental Biology 206, 2393–2397. Chittka, L., Beier, W., Hertel, H., Steinmann, E., Menzel, R., 1992. Opponent colour coding is a universal strategy to evaluate the photoreceptor inputs in hymentoptera. Journal of Comparative Physiology A 170, 545–563. De Valois, R.L., 1973. Central mechanisms of colour vision. In: Jung, R. (Ed.), Handbook of Sensory Physiology, vol. 7/3A. Springer, Berlin, pp. 209–253. Dyer, A.G., Chittka, L., 2004. Biological significance of discriminating between similar colours in spectrally variable illumination: bumblebees as a study case. Journal of Comparative Physiology A 190, 105–114. Ehmer, B., Gronenberg, W., 2002. Segregation of visual input to the mushroom bodies in the honeybee (Apis mellifera). Journal of Comparative Neurology 451, 362–373. Erber, J., Menzel, R., 1977. Visual interneurons in the median protocerebrum of the bee. Journal of Comparative Physiology A 121, 65–77. Von Frisch, K., 1914. Der farbensinn und formensinn der biene. Zoologische Fahrbu¨cher. Abteilung fu¨r allgemeine Zoologie und Physiologie der Tiere 35, 1–188. Von Frisch, K., 1967. The Dance Language and Orientation of Bees. Harvard University Press, Cambridge, MA. Fukushi, T., 1994. Colour perception of single and mixed monochromatic lights in the blowfly Lucilia cuprina. Journal of Comparative Physiology A 175, 15–22. Gegenfurtner, K.R., 2003. Cortical mechanisms of colour vision. Nature Reviews Neuroscience 4, 563–572. Giger, A., Srinivasan, M.V., 1996. Pattern recognition in honey bees: chromatic properties of orientation analysis. Journal of Comparative Physiology A 178, 763–769. Goldsmith, T.H., Bernard, G.D., 1974. The visual system of insects (Chapter 5). In: Rockstein, M. (Ed.), Physiology of Insecta, vol. VII/3A. Academic Press, New York, pp. 175–271. Von Helversen, O., 1972. Zur spektralen unterschiedsempfindlichkeit der honigbiene. Journal of Comparative Physiology 80, 439–472. Hertel, H., 1980. Chromatic properties of identified interneurons in the optic lobes of the bee. Journal of Comparative Physiology A 137, 215–231. Hertel, H., Maronde, U., 1987. Processing of visual information in the honeybee brain. In: Menzel, R., Mercer, A. (Eds.), Neurobiology and Behaviour of Honeybees. Springer, Berlin, pp. 141–157. Kaiser, W., 1974. The spectral sensitivity of the honeybee’s optomotor walking response. Journal of Comparative Physiology 90, 405–408. Kaiser, W., 1975. The relationship between visual movement detection and colour vision in insects. In: Horridge, G.A. (Ed.), The Compound Eye and Vision of Insects. Clarendon Press, Oxford, pp. 359–377. Kelber, A., Pfaff, M., 1999. True colour vision in the orchard butterfly, Papillio aegeus. Naturwissenschaften 86, 221–224.
Kelber, A., Vorobyev, M., Osorio, D., 2003. Animal colour vision— behavioural tests and physiological concepts. Biological Reviews 78, 81–118. Kien, J., Menzel, R., 1977a. Chromatic properties of interneurons in the optic lobes of the bee. I. Broad band neurons. Journal of Comparative Physiology A 113, 17–34. Kien, J., Menzel, R., 1977b. Chromatic properties of interneurons in the optic lobes of the bee. II. Narrow band and colour opponent neurons. Journal of Comparative Physiology A 113, 35–53. Kinoshita, M., Shimada, N., Arikawa, K., 1999. Colour vision of the foraging swallowtail butterfly Papillio xuthus. Journal of Experimental Biololgy 202, 95–102. Kolb, G., Scherer, C., 1982. Experiments on wavelength specific behaviour of Pieris brassicae L. during drumming and egg-laying. Journal of Comparative Physiology A 149, 325–332. Lehrer, M., 1987. To be or not to be a colour-seeing bee. Israel Journal of Entomology XXI, 51–76. Marshall, N.J., Jones, J.P., Cronin, T.W., 1996. Behavioural evidence for colour vision in stomatopod crustaceans. Journal of Comparative Physiology A 179, 473–481. Menzel, R., 1979. Spectral sensitivity and colour vision in invertebrates. In: Autrum, H. (Ed.), Handbook of Sensory Physiology, vol. 7/6A. Springer, Berlin, pp. 503–580. Menzel, R., 1985. Colour pathways and colour vision in the honeybee. In: Ottoson, D., Zeki, S. (Eds.), Central and Peripheral Mechanisms of Colour Vision. Macmillan Press, London, pp. 211–233. Menzel, R., Backhaus, W., 1989. Colour vision in honey bees: Phenomena and physiological mechanisms. In: Stavenga, D.G., Hardie, R.C. (Eds.), Facets of Vision. Springer, Berlin, pp. 281–297. Menzel, R., Blakers, M., 1976. Colour receptors in the bee eye— Morphology and spectral sensitivity. Journal of Comparative Physiology A 108, 11–33. Menzel, R., Snyder, A.W., 1974. Polarized light detection in the bee, Apis mellifera. Journal of Comparative Physiology 88, 247–270. Menzel, R., Ventura, D.F., Hertel, H., de Souza, J.M., Greggers, U., 1986. Spectral sensitivity of photoreceptors in insect compound eyes: comparison of species and methods. Journal of Comparative Physiology A 158, 165–177. Mo¨ller, R., 2002. Insect could exploit UV–green contrast for landmark navigation. Journal of Theoretical Biology 214, 619–631. Nathans, J., 1989. The genes for colour vision. Scientific American 260, 28–35. Neumeyer, C., 1980. Simultaneous colour contrast in the honeybee. Journal of Comparative Physiology A 139, 165–176. Neumeyer, C., 1981. Chromatic adaptation in the honeybee: successive colour contrast and colour constancy. Journal of Comparative Physiology A 144, 543–553. Osorio, D., Vorobyev, M., 1996. Colour vision as an adaptation to frugivory in primates. Proceedings of the Royal Society of London Series B 263, 593–599. Peitsch, D., Fietz, A., Hertel, H., de Souza, J., Ventura, D.F., Menzel, R., 1992. The spectral input systems of hymenopteran insects and their receptor-based colour vision. Journal of Comparative Physiology A 170, 23–40. Riehle, A., 1981. Colour opponent neurons of the honeybee in a heterochromatic flicker test. Journal of Comparative Physiology A 142, 81–88. Srinivasan, M.V., 1994. Visual discrimination of pattern orientation by honey bees: performance and implications for ‘cortical’ processing. Philosophical Transactions of the Royal Society of London Series B 343, 199–210. Strausfeld, N.J., Lee, J.-K., 1991. Neuronal basis for parallel visual processing in the fly. Visual Neuroscience 7, 13–33.
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925 Troje, N., 1993. Spectral categories in the learning behaviour of blowflies. Zeitschrift fu¨r Naturforschung 48c, 96–104. Werner, A., Menzel, R., Wehrhahn, C., 1988. Colour constancy in the honeybee. Journal of Neuroscience 8, 156–159. Zhang, S., Srinivasan, M.V., 1993. Behavioural evidence for parallel information processing in the visual system of
925
insects. Japanese Journal of Physiology 43 (Suppl. 1), S247–S258. Zrenner, E., Abramov, I., Akita, M., Cowey, A., Livingstone, M., Valberg, A., 1990. Colour perception: retina to cortex. In: Spillmann, L., Werner, J.S. (Eds.), Visual Perception: The NeuroPhysiological Foundations. Academic Press, San Diego, pp. 163–204.
Journal of Insect Physiology 50 (2004) 913–925 www.elsevier.com/locate/ip
Patterns of chromatic information processing in the lobula of the honeybee, Apis mellifera L. En-Cheng Yang, Hsiao-Chun Lin, Yu-Shan Hung a
Department of Entomology, National Chung Hsing University, 250 Kuo Kuang Rd., Taichung 402, Taiwan Received 14 March 2004; received in revised form 24 June 2004; accepted 28 June 2004
Abstract The honeybee, Apis mellifera L., is one of the living creatures that has its colour vision proven through behavioural tests. Previous studies of honeybee colour vision has emphasized the relationship between the spectral sensitivities of photoreceptors and colour discrimination behaviour. The current understanding of the neural mechanisms of bee colour vision is, however, rather limited. The present study surveyed the patterns of chromatic information processing of visual neurons in the lobula of the honeybee, using intracellular recording stimulated by three light-emitting diodes, whose emission spectra approximately match the spectral sensitivity peaks of the honeybee. The recorded visual neurons can be divided into two groups: non-colour opponent cells and colour opponent cells. The non-colour opponent cells comprise six types of broad-band neurons and four response types of narrow-band neurons. The former might detect brightness of the environment or function as chromatic input channels, and the latter might supply specific chromatic input. Amongst the colour opponent cells, the principal neural mechanism of colour vision, eight response types were recorded. The receptive fields of these neurons were not centre surround as observed in primates. Some recorded neurons with tonic post-stimulus responses were observed, however, suggesting temporal defined spectral opponency may be part of the colour-coding mechanisms. r 2004 Elsevier Ltd. All rights reserved. Keywords: Colour vision; Colour coding; Colour opponency; Honeybee; Apis mellifera
1. Introduction Colour is a meaningful signal for many animals, both vertebrate and invertebrate. Because colour discrimination is not present in all aspects of visual orientation (Lehrer, 1987; Zhang and Srinivasan, 1993), and because of the difficulty of behavioural training, so far, the only a few invertebrate species, e.g. bees (Frisch, 1914) and wasps (Chittka et al., 1992), flies (Troje, 1993; Fukushi, 1994), butterfly (for example, Kolb and Scherer, 1982; Kelber and Pfaff, 1999; Kinoshita et al., 1999) and stomatopods (Marshall et al., 1996), have been studied sufficiently carefully to draw the conclusion that they have true colour vision (rev. Autrum and Corresponding author. Tel./fax: +886-4-22856079
E-mail address: [email protected] (E.-C. Yang). 0022-1910/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.jinsphys.2004.06.010
Thomas, 1973; Goldsmith and Bernard, 1974; Menzel, 1979; Barth, 1985; Menzel and Backhaus, 1989; Kelber et al., 2003). Thanks to the successes of intracellular recording and behavioural reinforcement training, it has been convincingly demonstrated that the honeybee possesses a trichromatic visual system with UV (335 nm), blue (435 nm) and green (540 nm) receptors (Autrum and von Zwehl, 1964; Menzel and Blakers, 1976; Menzel et al., 1986; Peitsch et al., 1992). Behavioural investigations of colour vision in the honeybee have been carried out intensively since Karl von Frisch first demonstrated that bees see colour objects as a visual quality different from brightness, at least in the behavioural contexts of feeding and homing (von Frisch, 1914; rev. von Frisch, 1967). Behavioural experiments also demonstrated that, like humans, bees are able to correctly identify the colours of objects in
ARTICLE IN PRESS 914
E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
spite of dramatic changes in the spectral content of the illumination, that is to say they exhibit colour constancy (Neumeyer, 1981; Werner et al., 1988). It has also been demonstrated that the hue perceived by bees when viewing a given colour stimulus is altered after adaptation to a different colour stimulus. This phenomenon is known as successive colour contrast (Neumeyer, 1981; Dyer and Chittka, 2004). Colour perception is also influenced by the presence of other colours in the surround, an effect described as simultaneous colour contrast by Neumeyer (1980). Evidence from behavioural experiments has demonstrated that some behaviours related to particular visual functions is driven by a particular wavelength-specific neural pathway due to one of the three types of receptors. For example, honeybees use their UV receptors to detect polarized light and to determine orientation for navigation (Menzel and Snyder, 1974), and they use their green receptors for optomotor response (Kaiser, 1974, 1975) and for flight distance estimation (Chittka and Tautz, 2003). Studies on the form vision of the honeybee also show that the green receptors are the dominant signal source in the orientation-sensitive neurons (Srinivasan, 1994; Giger and Srinivasan, 1996). In contrast to the abundant information on behavioural experiments, physiological studies of honeybee colour vision are relatively few. Kien and Menzel (1977a, b) first reported three classes of colour-coding neurons in the optic lobes (mainly in the medulla and a few in the lobula) of the honeybee: broad-band neurons, narrow-band neurons and colour opponent neurons. The broad-band neurons can respond to light from across the broader spectrum, suggesting that they receive inputs from two or three receptor types. The narrow-band neurons are only sensitive to wavelengths within the spectral sensitivity of one photoreceptor type. The colour opponent neurons are excited by some wavelengths but are inhibited by others. From earlier studies in vertebrate visual systems we know that colour opponency is one of the most important neural mechanisms for processing colour information (De Valois, 1973). Previous studies of the optic lobes in the honeybee also showed similar colour opponent response patterns, however, the receptive fields of the recorded colour opponent neurons were not spatially antagonistic, i.e. did not have a centre-surround configuration, as some of those of vertebrates (Kien and Menzel, 1977a, b). In addition, by means of intracellular recording and staining, Hertel (1980) reported the response tunings of colour-coding neurons in the medulla and lobula. By using a heterochromatic flicker test for the colour opponent neurons of the lobula, Riehle (1981) provided neurophysiological evidence for successive colour contrast in bees.
Neural mechanisms have been an historic focus in primate colour vision systems. In invertebrates, most of the early studies of colour vision emphasized colour discrimination behaviour and the interaction between different spectral types of photoreceptors, while nowadays the trend is to focus on higher order neural mechanisms (Kelber et al., 2003). To study the neural mechanisms of the colour vision of a honeybee, it is important to see how the spectral signals are processed in the visual neurons. Only by identifying the response patterns of the light-sensitive neurons will we then be able to further investigate the roles each type of neuron plays in colour processing. In the present study we surveyed the spectral response patterns of the honeybee’s visual neurons with conventional intracellular recordings in the lobula. Together with previous studies, our results show that in total 19 possible combinations of spectral signals from the three receptors can be recorded from the light-sensitive neurons in the lobula.
2. Materials and Methods 2.1. Animals Experiments were carried out on adult honeybee workers (Apis mellifera L.) which were reared on the campus of the National Chung Hsing University in Taiwan. These honeybees were caught at the hive entrance while they were just returning to the hive. All honeybees were used immediately after capture. The honeybees were kept at room temperature (25721C) to sustain their physiological condition. 2.2. Preparation The honeybees were anaesthetized by placing them in a refrigerator at 41C for about 5 min. The thorax was then mounted on a plastic plate, which was fixed to a brass pedestal, with a beeswax/resin (3:1) mixture, so that the head and thorax were rigidly secured. The abdomen was free to perform ventilatory movements. The legs were amputated, and the wings were waxed together. The vertex of the head capsule between the left compound eye and the ocellus was removed to expose the optic lobe to admit recording electrodes. 2.3. Electrophysiological recording Electrophysiological recording was performed in a dark room to keep the honeybees dark-adapted. Microelectrodes, made of aluminosilicate glass capillaries (AF100-68-10, Sutter Instruments) and pulled on a micropipette puller (Flaming/Brown P-97, Shutter Instruments), had a resistance of between 70 and 90 MO when filled with 2 M potassium acetate and measured in
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
the brain tissue. The microelectrode was lowered vertically by a micromanipulator (MWS-32, Narishige) to penetrate the lobula. The manipulation was observed via a stereomicroscope (GZ6, Leica) to ensure that the microelectrode was accurately inserted into the lobula. The brass pedestal, with the honeybee mounted, and the micromanipulator were secured onto a vibration isolating workstation (LW3030-OPT, Newport) by means of magnetic bases. The indifferent electrode was a silver wire inserted into the thorax. The electrophysiological signal was amplified 10 (Neuroprobe, Model 1600, AM Systems). The signal was then displayed on an oscilloscope (GOS-620FG, GW-Instek), simultaneously acquired by a data-acquisition system (PCI-6024E, National Instruments) and stored in a personal computer for further processing. The control program for the acquisition system and the program for analysing the recorded data were both written in LabView (National Instruments). The recorded waveforms were digitized at sampling rate of 5 kHz. 2.4. Visual stimulus A conventional flash method was used to stimulate the recorded visual neurons. Three light-emitting diodes (LEDs) were used as the light source. The emission spectra of the three LEDs are shown in Fig. 1, and their peak wavelengths are: 370 nm (UV), 435 nm (blue) and 560 nm (green). These peak wavelengths were chosen according to match as closely as possible the spectral sensitivity of three types of honeybee photoreceptors (Peitsch et al., 1992). A multi-channel controller provided suitable voltages to drive each of the three LEDs. The intensities of the peak wavelengths of the UV, blue and green LEDs were 4.44 1013, 8.58 1013,
Fig. 1. Relative spectral radiant power distribution of the three LEDs used as visual stimulus in the experiment. The wavelengths of peak emission are at 370, 435 and 560 nm.
915
and 2.10 1013 photons/cm2/s, respectively. The light source was mounted on a Cardan arm perimeter, which could be accurately positioned at the point of maximum sensitivity of the recorded cell. Because only one of the three LEDs could be mounted on the Cardan arm, the LED was replaced when a colour stimulation change was needed. The angles subtended at the eye of the three light sources were: 3.31 for UV and 3.51 for both blue and green. The stimulating flash lasted for 1 s, and the interval between flashes was 10 s. Both the duration and the interval were controlled with an isolated pulse stimulator (Model 2100, A-M Systems). 2.5. Calibration A miniature fibre optic spectrometer (S2000, Ocean Optics) was used to determine the emission spectra of the LEDs. A radiometer (S370, UDT Instruments) with detectors (Model 222, 260, UDT Instruments) was employed to measure the light flux at the position of the eye. The energy measured at different monochromatic wavelengths was transformed into photons/cm2/s.
3. Results 3.1. General response properties to flash The spontaneous activity frequency and the amplitude of the recorded action potentials was within the range 0–80 Hz and 5–60 mV, respectively. Only those cells with spike amplitudes of more than 10 mV and basalline shifting of less than 10 mV were analysed in this report. More than 150 cells were recorded from the lobula. Among them, 97 cells were sensitive to flash stimuli. The receptive fields of the 97 cells were large (X601), and no centre-surround configurations were found. Kien and Menzel (1977a) divided the flashresponse profiles of the high-order visual neuron into three stages: on-effect, sustained response and off-effect. Both on- and off-effects are phasic, and the sustained response is tonic. Most of the 97 cells showed excitatory (67 cells) or inhibitory (13 cells) phasic responses, and only 17 cells responded tonically. The phasic response could be related to enhancing colour contrast, while the tonic response is presumed to be responsible for the sustained property of colour stimuli (Hertel and Maronde, 1987). Therefore, the analysis of neuronal responses to flashes is focused on the tonic response, and consequently the 34 cells showing only a phasic response are not considered in this report. Based on the tonic responses to different colour flash stimuli, the recorded neurons can be divided into two groups: non-colour opponent cells (45 cells) and colour opponent cells (10 cells).
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
916
3.2. Non-colour opponent cells We used the nomenclature of Kien and Menzel (1977a, b) to describe the 45 non-colour opponent cells. Thus, the 33 that cells were sensitive to more than one colour flash stimulus are referred to here as broad-band neurons. The remaining 12 cells were sensitive to one colour flash stimulus only and so are referred to as narrow-band neurons. Based on their response patterns, the broad-band neurons can be further categorized as:
Table 1 Non-colour opponent neural types, response patterns and numbers of colour-coding neurons recorded from the lobula of the honeybee Neural types Broad band neurons Flash-sensitive All excitatory All inhibitory Dual excitatory Dual inhibitory Narrow-band neurons Single excitatory Single inhibitory
Response patterns
No. of recording
Only phasic response UV+/B+/G+ UV /B /G B+/G+ UV /B B /G
34 10 8 5 8 2
UV+ B+ UV B
3 3 2 4
all excitatory neurons, all inhibitory neuron dual excitatory neuron, and dual inhibitory neuron (Table 1). Among the 33 broad-band neurons, 18 cells were recorded as sensitive to all three colour flash stimuli, including the ‘all excitatory’ neuron (UV+/B+/ G+, Fig. 2a) and the ‘all inhibitory neuron’ (UV /B / G , Fig. 2b). These two types of response patterns have been previously reported (Kien and Menzel, 1977a). The other 15 cells respond to two of the three colours: one type is the dual excitatory neurons (B+/G+, Fig. 3a) and these are also two types of inhibitory neurons (UV /B , Fig. 3b; B /G , Fig. 3c). These types of neurons have not been previously reported. Neither the response pattern of UV+/G+nor UV /G was recorded in our study. Kien and Menzel (1977a) once reported a neuron with a response pattern of UV+/ G+, and claimed that a neuron receiving inputs with the same sign from two non-overlapping receptors could in no way discriminate colour. It could, however, feed into a spectrally antagonistic neuron afterwards. The narrow-band neurons are categorized as: single excitatory neuron and single inhibitory neuron. Both UV-sensitive (UV+, Fig. 4a; UV , Fig. 4c) and blue-sensitive (B+, Fig. 4b; B , Fig. 4d) neurons were recorded; however, no green-sensitive neurons (neither G+ nor G ) was found. Green sensitive narrow-band neurons were previously reported (G+, Kien and Menzel, 1977a), but other types of
Fig. 2. The response patterns of flash-sensitive neurons elicited with three spectral light stimuli (UV, blue and green). (a) A neuron with phasic excitatory responses while stimulated with light on and off. (b) A different neuron showing phasic responses to light on and off. The horizontal bar under each response depicts the duration of a 1 s flash stimulus; the vertical bar at the right-hand side depicts response amplitude of 10 mV. Response histograms to 10 stimulations are shown below the response waveforms.
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
917
Fig. 3. The response patterns of broad-band neurons elicited with three spectral light stimuli (UV, blue and green). Dual excitatory neuron: (a) B+/ G+; dual inhibitory neurons: (b) UV /B , (c) B /G . The scales of the horizontal and vertical bars are the same as the depictions in Fig. 2.
narrow-band neurons (UV-and blue-sensitive only) are reported here for the first time. Numbers of recorded cells for each type of non-colour opponent neurons are listed in Table 1. 3.3. Colour opponent cells A total of 10 recorded cells with seven types of response patterns are colour opponent cells. Based on the opponent patterns these cells can be further categorized as: single opponent neuron, dual excitatory opponent neuron and dual inhibitory opponent neuron (Table 2). For the single opponent neurons, only the neurons inhibited by short wavelengths but excited by long wavelengths were recorded: UV /B+ (Fig. 5a), UV /G+ (Fig. 5b) and B /G+ (Fig. 5c); all the counter parts of the three opponent types, UV+/B , UV+/G and B+/G , were not found. Among the
three types of single opponent cells found, only the UV /G+ was previously reported (Hertel and Maronde, 1987). For the dual opponent neurons, three types of dual excitatory opponent neurons, UV+/B+/G (Fig. 6a), UV+/B /G+ (Fig. 6b) and UV /B+/G+ (Fig. 6c), and one type of dual inhibitory neuron, UV /B+/G (Fig. 7), was recorded. Previous studies on bee colour opponent cells indicated that UV+/B / G+ and UV+/B /G (and their counter parts UV /B+/G and UV /B+/G+) are the two main classes of colour opponent cells in the bee brain (Kien and Menzel, 1977b; Hertel, 1980; Menzel and Backhaus, 1989). Together with our present study only the counterpart of UV+/B+/G (i.e. UV /B /G+) was not found so far. Numbers of recorded cells for each type of colour opponent neurons are listed in Table 2.
ARTICLE IN PRESS 918
E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
Fig. 4. The response patterns of narrow-band neurons elicited with three spectral light stimuli (UV, blue and green). UV-sensitive neurons: (a) UV+, (c) UV ; blue sensitive neurons: (b) B+, (d) B . The scales of the horizontal and vertical bars are the same as the depictions in Fig. 2.
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925 Table 2 Colour opponent neural types, response patterns and numbers of colour-coding neurons recorded from the lobula of the honeybee Neural types Colour opponent neurons Single opponent
Dual excitatory opponent
Dual inhibitory opponent
Response patterns
No. of recording
UV /B+ UV /G+ B /G+ UV+/B+/G UV+/B /G+ UV /B+/G+ UV /B+/G
2 1 1 1 1 3 1
919
3.4. Post-stimulus response patterns During our recording some cells responded following the end of the colour flash. Two of six broad-band neurons are shown as examples in Fig. 8. The poststimulus responses were either excitatory (Fig. 8a) or inhibitory (Fig. 8b). The post-stimulus response patterns were not only found in non-colour opponent cells but also in colour opponent cells. Two colour opponent cells, as per examples shown in Figs. 6b and 7, exhibit strong post-stimulus responses to the three colour stimuli. Some colour opponent cells showed strong excitatory post-stimulus responses to the UV, blue and
Fig. 5. The response patterns of single opponent neurons elicited with three spectral light stimuli (UV, blue and green): (a) UV /B+; (b) UV /G+ and (c) B /G+. The scales of the horizontal and vertical bars are the same as the depictions in Fig. 2.
ARTICLE IN PRESS 920
E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
Fig. 6. The response patterns of three types of dual excitatory opponent neurons: (a) UV+/B+/G , (b) UV+/B /G+, and (c) UV /B+/G+. The scales of the horizontal and vertical bars are the same as the depictions in Fig. 2. Note that (b) shows strong post-stimulus responses to the three spectral stimuli: UV+/B+/G+.
Fig. 7. The response pattern of one type of dual inhibitory opponent neurons: UV /B+/G . The scales of the horizontal and vertical bars are the same as the depictions in Fig. 2. Note that the responses of this neuron shows strong post-stimulus responses to the three spectral stimuli: UV+/B /G+.
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
921
Fig. 8. The response patterns of two broad-band neurons with post-stimulus response elicited with three spectral light stimuli (UV, blue and green). (a) Tonic flash responses (UV+/G+), excitatory post-stimulus responses (UV+/B+/G+); (b) tonic flash responses (UV /B /G ), inhibitory post-stimulus responses (UV+/B /G+). The scales of the horizontal and vertical bars are the same as the depictions in Fig. 2.
green lights (Fig. 6b), and some even showed antagonistical responses. For example, the post-stimulus response pattern of the colour opponent cell shown in Fig. 7 was excited by UV and green lights but inhibited by blue, making the colour opponency of the poststimulus response UV+/B /G+, which is opposite to its tonic response patterns (UV /B+/G ).
4. Discussion 4.1. Response patterns of colour-coding neurons In the honeybee lobula, visual neurons can be divided into two groups according to whether or not they generate spontaneous responses (Hertel, 1980). Erber and Menzel (1977) suggested that the responses of highorder visual neurons elicited by different colour stimuli were similar, except for the frequency of spontaneous responses. In this present study all the 97 recorded neurons generated spontaneous responses, most of them (69%) with an on- or off-transient response when stimulated by flashed light. This result is similar to a previous study by Hertel (1980), in which a total of 88 visual neurons were recorded, and 52% of them gave
on-transient responses. In contrast to the transient response, the colour-coding neurons, which are of the most interest to us, should be able to sustain their responses to a prolonged colour stimulus. About 57% (55 neurons) in our recorded neurons had a tonic response to colour flashes, some of them providing both transient and tonic responses when stimulated by a flash. Previous recordings by Hertel (1980) showed that only 16% of recorded neurons had tonic responses. Since our recordings were confined to the lobula, and most recording sites of the previous study by Hertel (1980) were in the medulla, and only few cells were recorded from the lobula, this might explain the difference in the tonic response patterns of colour-coding neurons between the two studies. Table 3 shows the comparison of colour-coding neurons between previous studies and our results. Sixteen response patterns were recorded, of which seven of them are new discoveries. Together with the 12 response patterns found by previous studies (Kien and Menzel, 1977a, b; Hertel and Maronde, 1987), a total of 19 response patterns, distributed among all the categorized neural types (Tables 1 and 2), were observed. Although seven response patterns were still not revealed,
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
922
Table 3 The response patterns of recorded colour-coding neurons in the optic lobes of the honeybee, reported from this present study and from previous studies Response patterns Broad band UV+/B+/G+ UV /B /G UV+/B+a UV+/G+ B+/G+ UV /B UV /G a B /G Narrow band UV+ B+ G+ UV B G a Colour opponent UV+/B a UV /B+ UV+/G a UV /G+ B+/G a B /G+ UV+/B+/G UV+/B /G+ UV+/B /G UV /B /G+a UV /B+/G UV /B+/G+
Recorded neurons
Previous studies
— —
—(a) —(a) —(a)
— — — — — — —
—(b) —(b) —(b) —(b)
— — — — —
— —
—(c)
—(b) —(b) —(b) —(b)
a The response patterns have not been found yet and are only hypothesized. (a) Kien and Menzel (1977a); (b) Kien and Menzel (1977b) and (c) Hertel and Maronde (1987).
a close examination of these missing patterns (e.g. UV / B /G+) shows that their counter parts (e.g. UV+/B+/ G ) all were recorded (Table 3). 4.2. Neural mechanisms for honeybee colour vision Menzel (1979, 1985) proposed roles that each type of response pattern may play in processing the colour signals of insect vision. He suggested that the broadband neurons are designated as luminosity-type neurons, which are not involved in colour vision but are sensitive to the light intensity of the environment, and that the narrow-band neurons and colour opponent cells are related to colour vision and extract colour signals with different response patterns. In primates, the analogues of the all excitatory (UV+/B+/G+) and all inhibitory (UV /B /G ) neurons were suggested as to process signals for achromatic sensation (Zrenner et al., 1990). Kien and Menzel (1977a) presented three
types of broad-band neurons, in addition to UV+/B+/ G+ and UV /B /G , the response pattern of UV+/ G+ providing a type of neuron that is unable to distinguish UV from green. Our results show that the neurons recorded with dual inputs were from UV/blue or blue/green receptors. These dual-input broad-band neurons may not only be related to sensing luminosity but may, together with other colour-coding neurons, form input channels to higher-order neurons showing colour opponency. Narrow-band neurons are sensitive to the light stimulus within a restricted range of the spectrum, and their spectral sensitivities are usually similar to some photoreceptor type (Menzel, 1979). However, the spectral sensitivities of some narrow-band neurons recorded by Kien and Menzel (1977b) were narrower when compared to those of the photoreceptors. Although we did not measure the spectral sensitivity of our recorded neurons, our results show that only UVand blue-sensitive neurons were recorded, suggesting these neurons received inputs from UV and blue receptors only. The reason for no green-sensitive neuron being recorded could be due to the difference in recording sites. Our recording microelectrode was inserted vertically from the dorsal part of lobula, while Kien and Menzel (1977b) mainly recorded from the ventral part of the optic lobes. Because there are three main neural tracts for projections from the optic lobes to the protocerebrum in the visual system of a honeybee (Ehmer and Gronenberg, 2002), it is possible that the UV and blue narrow-band neurons might be in a tract different from the green-sensitive neurons. Neuroanatomical study of the parallel visual pathways in Diptera indicates that the retinotopic columns in the lamina comprise two pathways: the spectrally insensitive channel (R1-6/L1, 2, green sensitive) and the trichromatic or tetrachromatic channel (R7, 8, L3). The two channels separately supply two subsets of retinopotic neurons throughout the entire visual system according to their relationships with retina/lamina afferents (Strausfeld and Lee, 1991). To verify the separated channels in the visual system of the honeybee, further recordings with dye injection to reveal the neural profiles are necessary. It must be noted that the colour opponent cells shown in Kien and Menzel (1977b) were recorded from the medulla, rather than the lobula. Thus they suggested that the medulla was the region of the optic lobe in which colour signals were processed. Based on the two colour opponent types only, two counterparts for each type (type A: UV /B+/G+ and UV+/B /G ; type B: UV /B+/G and UV+/B /G+), Backhaus (1991) was able to construct a colour opponent coding model to explain the results obtained from the colour discrimination behaviour of the honeybee (von Helversen, 1972). Our results show that the colour
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
opponent cells are various, including the single opponent cells and the dual opponent cells (Table 2). Since the counterparts of all the missing types were recorded, it is very likely that all the 12 assumed types of colour opponent cells can be recorded from the lobula, although one certainly does not need all the types and further recordings are necessary to verify this point. One might wonder why so many different types of colour opponent cells exist in honeybees when only two are necessary to explain colour discrimination. It is reasonable to hypothesize that the single opponent cells may form the inputs for the three types of dual opponent cells, including a third pair (i.e. type C: UV+/B+/G and UV /B /G+). In addition, some possible functions related to particular visual channels have been identified by behavioural experiments, and our recorded cell types might be involved. For instance, a UV /G+ neuron and its counterpart might be part of a visual channel that provides UV–green contrast in order to discriminate an object against the background of a bright sky; such a channel might be employed when a bee uses landmarks to orient (Mo¨ller, 2002). 4.3. Colour opponent cells of primates and honeybees On the photoreceptor level, both primates and honeybees have three spectral types of photoreceptors, and thus both systems for colour vision are based on the theory of trichromacy. On the neural level, both primates and honeybees have colour opponent coding neurons, and therefore they can transfer the perceived spectral signals to colour signals. With colour vision, primates and bees are able to perform colour constancy and simultaneous colour contrast in behavioural experiments. Although trichromatic systems offer the animals to be able to discriminate colours, at the level of neural mechanisms for colour coding, the receptive fields of colour opponent cells of honeybees are different from those of primates. In contrast to the centre-surround receptive fields observed from primate visual neurons, to date no such receptive field is found from any recorded colour opponent cell of honeybees. Without the centresurround configuration, the colour opponent cells of honeybees would not be able to compare the contrast of neighbouring colours at the same time. Menzel and Backhaus (1989) proposed that simultaneous colour contrast might only exist for particular patterns of colour arrangements. Other colour contrast phenomena may be based solely on successive contrast effects, i.e. based on temporary sampling of contrast values. To determine the colour of an object independently of the chromatic illumination, the nervous system of honeybee must compare the wavelength composition of light reflected by an object with respect to the wavelength composition reflected from the surrounding
923
area. The neurons with post-stimulus response patterns reported in this present study are probably related to this hypothesis, and may provide the neural basis for the honeybee to compute the colour contrast when scanning over areas (Menzel and Backhaus, 1989). For the variety of response patterns, in contrast to only six patterns being found in the primates visual system: R+/G (R /G+), G+/R (G /R+) and B+/R G (B /R+/G+), and their counter parts (Zrenner et al., 1990), there are at least eight patterns in the honeybee’s visual system. This difference could be attributed to the evolutionary background of the two trichromatic systems. Whereas the rods and the shortwavelength (S) cones are phylogenetically older, the long (L) and middle-wavelength (M) cones evolved from a common ancestoral pigment only about 35 million years ago (Nathans, 1989; Osorio and Vorobyev, 1996; Bowmaker, 1998; Gegenfurtner, 2003). It has been known that primates have more red and green opponent cells than others (Zrenner et al., 1990). In contrast to the primates, the basic UV–blue–green trichromacy appears to date back to the Devonian ancestor of all pterygote insects, and each of the three visual pigments diverged very early in evolution (Briscoe and Chittka, 2001). The interaction of chromatic information from different types of photoreceptor in the insect visual system might have been developled since then. One might draw the inference that colour coding on the neuronal level might be more complex in bees than in humans (or primates). Previous behavioural work in bees by Backhaus and Menzel (Menzel and Backhaus, 1989; Backhaus, 1991) derived two axes for a psychophysical colour space which correspond to the colour opponent neurons reported by Kien and Menzel (1977b). While behavioural work seems to show that there is two dimensionality of colour space, colour coding on the neuronal level seems to be based on even more mechanisms than in humans.
Acknowledgements We would like to thank L. Chittka, D. Osorio and T. Maddess for reading and correcting our manuscript. Comments from two anonymous referees greatly improved our manuscript. This work was supported by a project grant (NSC 89-2313-13-005-166) from National Science Council, Taiwan, ROC.
References Autrum, H., Thomas, I., 1973. Comparative physiology of colour vision in animals. In: Jung, R. (Ed.), Handbook of Sensory Physiology, vol. 7/3A. Springer, Berlin, pp. 661–692.
ARTICLE IN PRESS 924
E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925
Autrum, H., von Zwehl, V., 1964. Die Spektrale Empfindlichkeit Einzelner Sehellen des Bienenauges. Zeitschrift fu¨r vergleichende Physiologie 48, 357–384. Backhaus, W., 1991. Colour opponent coding in the visual system of the honeybee. Vision Research 31, 1381–1397. Barth, F.G., 1985. Insect and Flowers—The Biology of a partnership (Translated by M. A. Biederman-Torson). Princeton University Press, Englewood Cliffs, NJ, p. 122. Bowmaker, J.K., 1998. Evolution of colour vision in vertebrates. Eye 12, 541–547. Briscoe, A.D., Chittka, L., 2001. The evolution of colour vision in insects. Annual Review of Entomology 46, 471–510. Chittka, L., Tautz, J., 2003. The spectral input to honeybee visual odometry. Journal of Experimental Biology 206, 2393–2397. Chittka, L., Beier, W., Hertel, H., Steinmann, E., Menzel, R., 1992. Opponent colour coding is a universal strategy to evaluate the photoreceptor inputs in hymentoptera. Journal of Comparative Physiology A 170, 545–563. De Valois, R.L., 1973. Central mechanisms of colour vision. In: Jung, R. (Ed.), Handbook of Sensory Physiology, vol. 7/3A. Springer, Berlin, pp. 209–253. Dyer, A.G., Chittka, L., 2004. Biological significance of discriminating between similar colours in spectrally variable illumination: bumblebees as a study case. Journal of Comparative Physiology A 190, 105–114. Ehmer, B., Gronenberg, W., 2002. Segregation of visual input to the mushroom bodies in the honeybee (Apis mellifera). Journal of Comparative Neurology 451, 362–373. Erber, J., Menzel, R., 1977. Visual interneurons in the median protocerebrum of the bee. Journal of Comparative Physiology A 121, 65–77. Von Frisch, K., 1914. Der farbensinn und formensinn der biene. Zoologische Fahrbu¨cher. Abteilung fu¨r allgemeine Zoologie und Physiologie der Tiere 35, 1–188. Von Frisch, K., 1967. The Dance Language and Orientation of Bees. Harvard University Press, Cambridge, MA. Fukushi, T., 1994. Colour perception of single and mixed monochromatic lights in the blowfly Lucilia cuprina. Journal of Comparative Physiology A 175, 15–22. Gegenfurtner, K.R., 2003. Cortical mechanisms of colour vision. Nature Reviews Neuroscience 4, 563–572. Giger, A., Srinivasan, M.V., 1996. Pattern recognition in honey bees: chromatic properties of orientation analysis. Journal of Comparative Physiology A 178, 763–769. Goldsmith, T.H., Bernard, G.D., 1974. The visual system of insects (Chapter 5). In: Rockstein, M. (Ed.), Physiology of Insecta, vol. VII/3A. Academic Press, New York, pp. 175–271. Von Helversen, O., 1972. Zur spektralen unterschiedsempfindlichkeit der honigbiene. Journal of Comparative Physiology 80, 439–472. Hertel, H., 1980. Chromatic properties of identified interneurons in the optic lobes of the bee. Journal of Comparative Physiology A 137, 215–231. Hertel, H., Maronde, U., 1987. Processing of visual information in the honeybee brain. In: Menzel, R., Mercer, A. (Eds.), Neurobiology and Behaviour of Honeybees. Springer, Berlin, pp. 141–157. Kaiser, W., 1974. The spectral sensitivity of the honeybee’s optomotor walking response. Journal of Comparative Physiology 90, 405–408. Kaiser, W., 1975. The relationship between visual movement detection and colour vision in insects. In: Horridge, G.A. (Ed.), The Compound Eye and Vision of Insects. Clarendon Press, Oxford, pp. 359–377. Kelber, A., Pfaff, M., 1999. True colour vision in the orchard butterfly, Papillio aegeus. Naturwissenschaften 86, 221–224.
Kelber, A., Vorobyev, M., Osorio, D., 2003. Animal colour vision— behavioural tests and physiological concepts. Biological Reviews 78, 81–118. Kien, J., Menzel, R., 1977a. Chromatic properties of interneurons in the optic lobes of the bee. I. Broad band neurons. Journal of Comparative Physiology A 113, 17–34. Kien, J., Menzel, R., 1977b. Chromatic properties of interneurons in the optic lobes of the bee. II. Narrow band and colour opponent neurons. Journal of Comparative Physiology A 113, 35–53. Kinoshita, M., Shimada, N., Arikawa, K., 1999. Colour vision of the foraging swallowtail butterfly Papillio xuthus. Journal of Experimental Biololgy 202, 95–102. Kolb, G., Scherer, C., 1982. Experiments on wavelength specific behaviour of Pieris brassicae L. during drumming and egg-laying. Journal of Comparative Physiology A 149, 325–332. Lehrer, M., 1987. To be or not to be a colour-seeing bee. Israel Journal of Entomology XXI, 51–76. Marshall, N.J., Jones, J.P., Cronin, T.W., 1996. Behavioural evidence for colour vision in stomatopod crustaceans. Journal of Comparative Physiology A 179, 473–481. Menzel, R., 1979. Spectral sensitivity and colour vision in invertebrates. In: Autrum, H. (Ed.), Handbook of Sensory Physiology, vol. 7/6A. Springer, Berlin, pp. 503–580. Menzel, R., 1985. Colour pathways and colour vision in the honeybee. In: Ottoson, D., Zeki, S. (Eds.), Central and Peripheral Mechanisms of Colour Vision. Macmillan Press, London, pp. 211–233. Menzel, R., Backhaus, W., 1989. Colour vision in honey bees: Phenomena and physiological mechanisms. In: Stavenga, D.G., Hardie, R.C. (Eds.), Facets of Vision. Springer, Berlin, pp. 281–297. Menzel, R., Blakers, M., 1976. Colour receptors in the bee eye— Morphology and spectral sensitivity. Journal of Comparative Physiology A 108, 11–33. Menzel, R., Snyder, A.W., 1974. Polarized light detection in the bee, Apis mellifera. Journal of Comparative Physiology 88, 247–270. Menzel, R., Ventura, D.F., Hertel, H., de Souza, J.M., Greggers, U., 1986. Spectral sensitivity of photoreceptors in insect compound eyes: comparison of species and methods. Journal of Comparative Physiology A 158, 165–177. Mo¨ller, R., 2002. Insect could exploit UV–green contrast for landmark navigation. Journal of Theoretical Biology 214, 619–631. Nathans, J., 1989. The genes for colour vision. Scientific American 260, 28–35. Neumeyer, C., 1980. Simultaneous colour contrast in the honeybee. Journal of Comparative Physiology A 139, 165–176. Neumeyer, C., 1981. Chromatic adaptation in the honeybee: successive colour contrast and colour constancy. Journal of Comparative Physiology A 144, 543–553. Osorio, D., Vorobyev, M., 1996. Colour vision as an adaptation to frugivory in primates. Proceedings of the Royal Society of London Series B 263, 593–599. Peitsch, D., Fietz, A., Hertel, H., de Souza, J., Ventura, D.F., Menzel, R., 1992. The spectral input systems of hymenopteran insects and their receptor-based colour vision. Journal of Comparative Physiology A 170, 23–40. Riehle, A., 1981. Colour opponent neurons of the honeybee in a heterochromatic flicker test. Journal of Comparative Physiology A 142, 81–88. Srinivasan, M.V., 1994. Visual discrimination of pattern orientation by honey bees: performance and implications for ‘cortical’ processing. Philosophical Transactions of the Royal Society of London Series B 343, 199–210. Strausfeld, N.J., Lee, J.-K., 1991. Neuronal basis for parallel visual processing in the fly. Visual Neuroscience 7, 13–33.
ARTICLE IN PRESS E.-C. Yang et al. / Journal of Insect Physiology 50 (2004) 913–925 Troje, N., 1993. Spectral categories in the learning behaviour of blowflies. Zeitschrift fu¨r Naturforschung 48c, 96–104. Werner, A., Menzel, R., Wehrhahn, C., 1988. Colour constancy in the honeybee. Journal of Neuroscience 8, 156–159. Zhang, S., Srinivasan, M.V., 1993. Behavioural evidence for parallel information processing in the visual system of
925
insects. Japanese Journal of Physiology 43 (Suppl. 1), S247–S258. Zrenner, E., Abramov, I., Akita, M., Cowey, A., Livingstone, M., Valberg, A., 1990. Colour perception: retina to cortex. In: Spillmann, L., Werner, J.S. (Eds.), Visual Perception: The NeuroPhysiological Foundations. Academic Press, San Diego, pp. 163–204.