Day-night changes in rhabdom size of compound eyes were investigated in three groups of crickets (Gryllus bimaculatus): nymphs and adult males and females. In both adults and nymphs, the rhabdoms were larger at night than during a day. In adults, the mean rhabdom occupation ratios (RORs) of ommatidial retinulae at midnight were about two times greater than the values at midday. This change contributes to control of the photon capture efficiency (PCE) of the eye according to photic environment. The RORs of adult males at midnight were higher than those of both adult females and nymphs. This suggests that the PCE of the compound eye of adult males is the greatest of all groups. Under constant darkness, day-night changes in ROR were detected only in adult males, but neither in adult females nor in nymphs. On the other hand, no day-night changes were detected in any experimental group under constant light. These results suggest that the change in rhabdom size between day and night is an adaptation to the photic environment that is controlled mainly by the light-dark (day-night) cycle. However, the change in male adults is induced by an endogenous circadian clock.
Several comparative studies have reported differences in the structure and function of the compound eyes of diurnal and nocturnal arthropods (Land, 1981; Eguchi, 1982; Laughlin and Weckström, 1993; Land et al., 1999). As an example of one of structural differences, the rhabdoms of nocturnal lepidopterans such as moths are much bigger than those of diurnal butterflies, presumably resulting in greater sensitivity to dim light at night (Eguchi, 1982). In fact, it has been demonstrated that the compound eyes of nocturnal species are much more sensitive to dim light than those of diurnal species (Eguchi and Horikoshi, 1984).
Circadian changes in the sensitivity of compound eyes were reported in several species, including adult crickets (Tomioka and Chiba, 1982b), cockroaches (Wills et al., 1986), sea slaters (Hariyama et al., 1986, 2001), grapsid crabs (Arikawa et al., 1987) and horseshoe crabs (Barlow et al., 1987). A rhabdom containing visual substance is the initial site of visual excitation in retinular cell in the compound eye; it is characterized by tightly packed tubular membranes called microvilli. The increase in photon capture efficiency (PCE), primarily depends on morphological factors such as larger rhabdom size, development of palisades surrounding rhabdoms in the photoreceptor cell cytoplasm (for light reflection to rhabdoms), and migration of screening and/or reflecting pigment granules in compound eyes according to light and dark adaptation (Horridge and Barnard, 1965; Williams, 1982; Arikawa et al., 1987; review by Meyer-Rochow, 1999). In particular, it can be said that the changes in rhabdom sizes influence PCE much greater than the effects of palisades and pigment granules in compound eyes. Because, a rhabdom, in general, has a property of waveguide of light, namely the light entered in a rhabdom propagates inside the rhabdom by repeating internal refelection in the rhabdom, which means small quantity of light may leak from a rhabdom (for review see Stavenga, 1979).
The cricket Gryllus bimaculatus is diurnally active in the nymphal stage, however, in the adult stage it becomes nocturnally active. Tomioka and Chiba (1982a) reported that a reversal of circadian rhythm occurs on 4 to 5 days after imaginal molt. It would be useful to know how the structure and function of compound eyes change before and after this reversal of circadian rhythm from diurnal nymph to nocturnal adult. First, the compound eyes of both the nymphal and adult stages are of apposition-type (Fig. 1), i.e., generally adapted to photopic vision despite a nocturnal life-style in the adult stage, as previously reported by Nilsson et al. (1987). Second, there are specialized ommatidia in the dorsal rim area (DRA) of the cricket eye (Fig. 1F, G). Functionally, the DRA is characterized by extremely high sensitivity to polarized light (Labhart et al., 1984). Morphologically, the DRA is clearly distinguished from the other main region by its characteristic structures, i.e., its flat corneal surface, (Ukhanov et al., 1996), larger rhabdoms, and absence of pigment granules in the cytoplasm of its photoreceptor cells (Burghause, 1979).
Hoff (1985) has reported that the cross-sectional areas of the rhabdoms of the compound eye increased from day to night by almost two-fold in the adult Gryllus bimaculatus, and therefore the capacity of photon absorption of the eyes also increased significantly at night. She concluded that this change in rhabdom size is suited to the cricket life-style, i.e., it lives in a wide range of light intensities.
In this study, we sought to clarify the changes in the structure of the compound eye between day and night in relation to the rhythm-reversal from nymph to adult. We focused on the rhabdom occupation ratios (RORs) of ommatidia as one of the major factors influencing eye PCE. RORs were measured separately in the DRA and in the central area of the compound eye, in both the nymphal and adult stages. We also compared PCE of compound eyes between day and night estimated by a theoretical equation based on morphological factors of the compound eyes (Land, 1981).
MATERIALS AND METHODS
Animals and light/dark regimes
Nymphs (in the last instar) and adult crickets (Gryllus bimaculatus) were used in the present study. They were obtained from a laboratory colony maintained under a 12:12 light/dark regime (light: 04:00–16:00, local time) at 27°C. Daylight fluorescent lamps were used for the light source, and light intensity was about 1000 lux. We put some tips of crumpled white paper in the cage as shelters for the crickets. The light intensity under these paper shelters was about 500 lux. The compound eyes were cut at midday (at 10:00) or at midnight (at 22:00). In addition, to clarify whether the structural changes were induced by an intrinsic biological circadian rhythm or not, some animals were moved into a constant dark or light box, and the eyes were fixed at their subjective midday (at 10:00) or subjective midnight (at 22:00) on second day in their constant regimes. For the experiments at midnight and under constant darkness, dissections were performed under a dim red light to prevent light adaptation of the eyes.
Compound eyes were prefixed in 2% glutaraldehyde and 2% paraformaldehyde in 0.1 M cacodylate buffer at pH 7.3 for 12–15 h at 4°C. For the prefixation of the eyes at midnight and under constant darkness, vials containing the eyes were placed in a dark room. After washing with the same buffer, the tissues were postfixed for 2–3 h in 2.0% OsO4 in the same buffer, dehydrated in an acetone series and then embedded in Epon. For light microscopy, semi-thin sections were cut with glass knives, stained with toluidine blue, and examined with a light microscope (Olympus BH-2). For electron microscopy, ultra-thin sections were cut with a diamond knife, double-stained with uranyl acetate and lead citrate for 5 min each, and then examined under a JEM 1200EX electron microscope.
Measurement of rhabdom occupation ratio (ROR) of an ommatidium
In order to examine the day-night change in rhabdom size, RORs were measured in light micrographs. Pictures of cross-sections containing at least 20 hexagonally arranged ommatidia were examined from the central region and from the DRA, at the level of the nucleus of a retinular cell. Rhabdom area and inter-ommatidial distance (center-to-center) were measured by an image analyzing system connected to a computer. Generally, shapes and sizes of retinular and pigment cells are not constant among ommatidia. Thus, in order to reduce measurement errors, the mean value of all inter-ommatidial distances for an eye was taken as the mean ommatidial diameter at the level of the nuclei of retinular cells for calculation of the ROR. A Mann-Whitney U-test was used for statistical analysis of the RORs between day and night in the same developmental stage and sex of crickets. In addition, a Kruskal-Wallis test and a Scheffe's test were used for multiple comparison among different groups under same light condition.
Calculation of sensitivity
Photon capture efficiency (PCE) of an ommatidium was calculated from Land's equation (Land, 1981):
Diameter of a corneal lens, cross-sectional rhabdom area and rhabdom length were measured in light micrographs. As for the focal length of an ommatidium, the distance between the inner cor-neal surface and the backfocal plane was measured following Ukhanov et al., (1996). Namely, the retinal tissue was removed from a freshly-cut eyecup by a thin strip of filter paper. The eyecup (composed of corneal lenses and crystalline cones) was rinsed with Ringer solution and the cleaned eyecup was placed on a coverslip so that the corneal surface was exposed to air. The preparation then was sealed in a moist chamber and placed on the stage of a light microscope (Olympus BH-2) with the outer surface of the cor-neal lens oriented toward the condenser lens of the microscope. The backfocal plane was found by a sharp image of a square shape drawn on a transparent sheet just above the light source of the microscope. The backfocal plane and morphological factors were measured in 10 ommatidia in the central region of 5 compound eyes, taken from both nymphs and adults, and their averaged values were used for the PCE calculation. A Mann-Whitney U-test was used for the statistical analysis between day and night in the same developmental stage and sex of the groups of crickets.
The structure of a compound eye
The central region of the compound eye
The compound eyes of both nymphs and adults are of the aposition type, and their ommatidial structures are basically the same as those shown diagrammatically in Fig. 1. The small compound eyes of the early nymphal stage gradually become larger as nymphs become adults. For example, rhabdoms of approximately 50–60 μm in length in the early nymphal stage become 200–250 μm long in adults. Typically, a long proximal process of a crystalline cone elon-gates deeply in the center of an ommatidial rhabdom (Figs. 1A–C and 2). One ommatidial retinula consists of 8 retinular cells, numbered 1-8 by Burghause (1979). The No. 1, 3, 5 and 7 cells have crystalline cone tracts on both sides at the distal layer (Fig. 1B). Microvilli of an ommatidial rhabdom are, in principle, oriented in two directions perpendicular to each other (Figs. 1 and 3). In the most distal layer, a rhabdom is composed of 4 rhabdomeres, numbered R1, R3, R5 and R7, where the microvilli of R1 and R5 are perpendicular to those of R3 and R7. This pattern of rhabdom formation is seen throughout almost its entire length (Fig. 1B-E). In the central and proximal layers, the microvilli of rhabdomeres from pairs R1, R8 (R8 only in the limited proximal layer) and R4, R5 are oriented in the same direction, while those from pairs R2, R3 and R6, R7 are oriented in a perpendicular direction (Fig. 1C-E).
The dorsal rim area of the compound eye
Ommatidial structure in the DRA is characterized by its flattened cornea; the rhabdom is 4–5 times bigger than that from the central region of the eye and there are no pigment granules in the retinular cells (Fig. 1F, G). An ommatidial retinula is composed of 8 retinular cells, same as in the central region of the eye. An ommatidial rhabdom is composed of 6 rhabdomeres (R1, R2, R5, R6, R7 and R8), but R3 and R4 do not contribute to the rhabdom formation (Figs. 1G and 4). As shown in Fig. 1G, microvilli from R1, R2, R5, R6 and R8 are arranged in the same direction, while those from R7 are perpendicular to those of the other cells. Functionally, it is well known that the DRA is characterized by a significantly high polarization sensitivity (Labhart et al., 1984).
Structural changes between day and night
We examined day-night structural changes in the compound eyes of four groups, male and female nymphs and adult males and females, under the different light and dark regimes mentioned in “Materials and methods”. In all cases, male nymphs did not display any detectable differences from female nymphs (data not shown). In the following sections, the results from male and female nymphs are combined and simply described as nymphs.
The central region of the compound eye
Fig. 5A shows the RORs in the central region of the compound eyes between midday and midnight under a 12:12 light:dark regime. The RORs at midnight (nymph 2.5%, adult male 4.1%, adult female 3.0%) were significantly higher than those at midday (nymph 1.5%, male adult 2.0%, female adult 1.1%), in both nymphal and adult stages (nymph, P < 0.01; adult male, P < 0.005; adult female, P < 0.05). The rhabdom of a dark-adapted eye at midnight was round in cross-section, surrounded by palisades, and there were few pigment granules around the rhabdom (Fig. 3A). On the other hand, the rhabdom cross-sections of light-adapted eyes at midday were square in shape, and the pigment granules aggregated around the rhabdom and the palisades disappeared (Fig. 3B). Moreover, many lamellated bodies (LBs) and characteristic structures of regularly arranged and wavy endoplasmic reticulum were seen in the cytoplasm of light-adapted retinular cells (Figs. 6 and 7). Such wavy endoplasmic reticulum was frequently observed by the side of LBs in retinular cells, and we could confirm that some of them were directly continuous with each other (Fig. 7B). In contrast to the pigment granules in the retinular cells, pigment granules in the pigment cells did not show notable migrations between day and night (data not shown).
At midnight, the ROR of adult male crickets (4.1%) was larger than that of nymphs (2.5%) (P < 0.05), and the ROR of adult female crickets (3.0%) did not show significant difference to that of nymphs (by Kruskal-Wallis and Scheffe's tests). This result suggests that, at night, the retinular cells of adult males have higher sensitivity than those of adult females and nymphs, even under the same light conditions.
The dorsal rim area of the compound eye
Fig. 5B shows the RORs of the DRAs of the compound eyes at midday and midnight. In contrast to the central region of the eye, the RORs of the DRAs of adults did not exhibit any significant differences between midday and midnight (male adult, 22.3∼24.4%; female adult, 16.2∼20.0%). In nymphs, strangely, the RORs at midday (21.0%) were significantly higher (17.5%) than at midnight (P < 0.05).
Changes in photon capture efficiency (PCE) between day and night
To compare the PCE of ommatidia using Land's equation (1981), we measured the necessary morphological factors of ommatidia in adult males, females and nymphs separately; the PCE of the compound eye was calculated for each group from the average values (Table 1). Cross-sectional rhabdom areas (S) increased from midday to midnight in all groups (nymphs, 14.3∼23.8 μm2, P < 0.005; adult males, 21.0∼43.0 μm2, P < 0.005; adult females, 12.1∼32.8 μm2, P < 0.05). However, the day-night differences were not statistically significant (P > 0.1) for other factors such as focal length (f) (63.9∼69.4 ìm), rhabdom length (x) (243.6∼291.4 μm) and diameter of a corneal lens (A) (34.7∼37.4 μm). As shown in Table 1, the PCEs of the compound eyes of all groups at night were approximately twice as those of the corresponding group during the day. Among the groups, the PCE of the compound eyes of adult males at midnight was the highest (8.8), while the daytime PCE of adult female compound eyes was the lowest value obtained (2.5).
Ommatidial structures and calculated photon capture efficiencies (PCEs) (see Materials and methods). A, diameter of a corneal lens; S, rhabdom cross-sectional area; x, rhabdom length; f, focal length of a corneal lens. All factors are averages from 10 ommatidia in 5 compound eyes each, and the PCEs are calculated from averaged values. The results of statistical comparison in an identical group between day and night, using a Mann-Whitney U-test, are shown with the significance level (P). NS, not significant.
Circadian structural changes
Under constant darkness, only the RORs of adult male crickets showed a significant difference between subjective day (4.4%) and subjective night (6.6%) (P < 0.05; Fig. 8A). This indicates that the changes in rhabdom size are controlled by an endogenous circadian clock, at least in adult male crickets. On the other hand, under constant light with 1000 lux, we could not detect any significant difference between subjective day (2.2∼3.0%) and subjective night (2.2∼3.0%) in any experimental group (Fig. 8B). The inhibitory effect of constant light on the increase of rhabdom size appears to be stronger than that of the endogenous circa-dian clock.
In contrast to the changes in rhabdom size, we could detect neither any circadian rhythm in palisade formation around the rhabdom nor migration of pigment granules in the retinular cell. Regardless of time of the day, many palisades were observed around the rhabdom and pigment granules were diffused in dark-adapted eye. While, palisades disappeared and pigment granules aggregated around the rhabdom in light-adapted eye (data not shown). These results indicate that the formation of palisades is solely caused by dark adaptation and aggregation of pigment granules by light adaptation in the cricket eye.
We have discovered several interesting diurnal features of cricket compound eye. (1) Under normal light/dark cyclic conditions, RORs increased significantly from day (light) to night (dark) in both nymphs and adult crickets, except for the DRAs of nymphs, which showed the reverse trend. (2) Other morphological factors that influenced the sensitivity of the eye, such as length of rhabdom, lens diameter, etc., all did not show significant change between day and night. Thus ROR is the major factor which influenced PCE of the eye of crickets. (3) Under constant darkness, the RORs of nymphs and female adults exhibited no change; only the RORs of adult males independently displayed a free-running circa-dian rhythm. (4) Under conditions of constant light, the RORs of all groups remained consistently at a low level, and did not display any circadian rhythm. These results represent the first report of day-night morphological changes in the compound eye of the cricket Gryllus bimaculatus by developmental stage and sex. Hoff (1985) has already reported results similar to (1) in adult crickets.
Changes in ommatidial structure and photon capture efficiency
The central region of the compound eye
The increase in rhabdom size from day to night has been reported previously in some other species, such as locusts Locusta migratoria (Williams, 1982), sea slaters Ligia exotica (Hariyama et al., 1986) and crabs Hemigrapsus sanguineus (Arikawa et al., 1987). In locusts (Williams, 1982) and crabs (Arikawa et al., 1987), the formation of palisades around rhabdoms at night has also been observed as in cricket eyes; these contribute to the heightened sensitivity of the eye at night. The change in rhabdom size is determined by a reciprocal relationship between the synthesis and degradation of the rhabdom membrane (Eguchi, 1999). Thus, if the synthesis of rhabdom membrane is greater than its degradation, the rhabdom increases in size, while if degradation dominates synthesis, its size decreases. The LBs and wavy endoplasmic reticulum in the cytoplasm of the cell body of a retinular cell are probably associated with the processes of membrane turnover of rhabdoms. Similar lamellar structures were reported in the compound eyes of horseshoe crabs Limulus (Chamberlain and Barlow, 1984), where they appeared during the process of membrane shedding caused by the light-onset or endogenous circadian clocks. Since many LBs (Fig. 6) and wavy endoplasmic reticulum (Fig. 7A) appeared at midday, this suggests that the cricket rhabdom membrane is also shed under daylight conditions.
The change in PCE of the cricket compound eye from day to night depends solely on the change in rhabdom size, because the other morphological factors exhibit no significant changes according to day and night. The PCE of the compound eyes of all groups of crickets examined in the present experiments ranged from a low of 2.5 (adult females at day) to a high of 8.8 (adult males at night). The value 2.7 for nymph compound eyes at daytime is very high for diurnal insects; Land (1981), for example, reported 0.32 for Apis workers. On the other hand, the 8.8 of adult males at night is much lower than the 82.8 of the typical nocturnal moth Ephestia (Land, 1981). This may reflect a difference in lifestyle, that is, visual information is more important for flying insects, such as moths, than for field insects, such as crickets. Nymphs shows the smallest difference in sensitivity between day and night because they are diurnally active, that is to say, they may not use their visual sense in dim light environments such as night.
The dorsal rim area of the compound eye
Hoff (1985) reported that rhabdom size in the DRA also changed according to day and night. Contrary to this observation, we could find no clear change in the RORs of ommatidia in the DRA of the adult eye. A possible cause for the lack of agreement could be attributed to differences in light intensity. Namely, in Hoff's experiments, crickets were put in a small cage so that they were directly exposed to light, while our crickets were put in a large cage with paper shelters (see Materials and methods). Under natural conditions in the field, it is known that both nymphs and adults avoid direct bright light and spend most of the day hidden under shaded shelters (Honegger and Campan, 1989). Therefore, it can be said that our experimental conditions were closer to the cricket's natural habitat. The fact that the change in ROR occurred in the central area but not in the DRA under the same conditions, suggests that stronger light may be necessary to induce shedding of rhabdom membrane in the DRA than is required in the central area of the eye. It is unclear why daytime rhabdom size in the DRA of nymphs was significantly larger than their nighttime rhabdom size.
The bigger rhabdom in the DRAs of the cricket's compound eye even in daytime could contribute to keeping relatively high PCE under light condition. Since the detection of polarized skylight in an ommatidium in DRA is attained through the polarization antagonism caused by e-vectors inputs from the mutually orthogonal oriented microvilli (Labhart and Mayer, 1999).
Effect of the circadian clock on rhabdom size
It is notable that only adult males showed a distinct circadian change in rhabdom size under conditions of constant darkness (Fig. 8A). The compound eyes of adult males were much more influenced by the endogenous circadian clock than the eyes of the other cricket groups in the present experiments. The fact that no clear change was observed in the ROR under constant light suggests that light reduced the rhabdom area regardless of the time of day. Such circadian changes in structure and function of compound eyes under constant dark or light conditions have been intensively studied in several other species. They have been studied morphologically in: horseshoe crabs, Chamberlain and Barlow (1984); crabs, Arikawa et al. (1987); and sea slaters, Hariyama et al. (1986). They also have been studied electrophysiologically in several species: including cockroaches, Wills et al. (1986); crickets, Tomioka and Chiba (1982b); and crabs, Arikawa et al. (1987).
In the cricket Gryllus bimaculatus, the optic laminamedulla complex is thought to be the site of the circadian clock (Tomioka and Chiba, 1986, 1992). It also has been reported that the amplitude of an electroretinogram (ERG) from an adult male cricket compound eye showed circadian rhythm under constant darkness; this rhythm was maintained even after the optic lamina-medulla complex and compound eye were neurally isolated from the central nervous system (Tomioka and Chiba, 1982b).
Reversal of circadian rhythm from nymph to adult stage
It is well known that crickets are diurnally active in the nymphal stage, but that they become nocturnal 4 to 5 days after the imaginal molt (Tomioka and Chiba, 1982a). Therefore, high sensitivity to light at night is not necessary for nymphs. We know that the ovipositional activity of mated females is significantly higher at night (Sugawara, 1997, 1998) and male calling song elicits strong locomotor activity in virgin females regardless of the time of a day (Loher, 1979). Virgin females walk around searching for a mate mainly by phonotaxs, but after mating, they devote themselves to laying eggs, with little locomotion. Thus, it is likely that visual information is less important for females. The present experiments revealed that the compound eyes of adult males exhibit higher PCE (8.8) at night than females' (6.4) and nymphs' (4.3) eyes at night. This fact may reflect that visual information is more important to males at night.