Appl. Entomol. Zool. 38 (3): 281–291 (2003)

Post-embryonic photoreceptor development and dark/light adaptation in the stick morosus (Phasmida, Phasmatidae)

V. Benno MEYER-ROCHOW1,* and Essi KESKINEN2 1 School of Engineering and Science, International University Bremen (IUB); D-28725 Bremen, Germany 2 Department of Biology, University of Oulu; P.O. Box 3000, SF-90014 Oulu, Finland (Received 28 October 2002; Accepted 7 March 2003)

Abstract The aims of this paper have been (a) to document postembryonic eye growth in the common laboratory stick insect and (b) to examine whether the capacity of the eye to adapt to changing light intensities varied with age. We found that number of facets, corneal thickness, ommatidial diameter, widths of cone and retinal layers, and rhabdom volume increased linearly. Interommatidial angles, pigment grain sizes, and microvillar diameters, how- ever, remained approximately the same. On the basis of the morphometric data we calculated that sensitivity in the adult stick insect eye was at least tenfold that of the eye of first instar nymphs and could show that adults would be able to perceive a similar amount of detail at considerably dimmer ambient light than the smaller individuals. In terms of resolving power this means that light- and dark-adapted first instar stick possess acceptance angles of 5.3° and 8°, respectively and that the corresponding figures for the adult eye are 4.7° and 7.3°. The bigger (and more light- sensitive) eye of fully grown C. morosus makes protection against radiation damage a more serious issue in the adult individual than the small first instar. The findings explain why smaller (ϭyounger) stick insects are less nocturnal than mature, fully-grown individuals. It is, thus, not surprising to see that dark/light adaptational photomechanical changes affecting pigment position and rhabdom widths, are more pronounced in the eyes of the adults.

Key words: Vision; compound eye; growth; optics; visual behaviour

down position. Observations like these, supple- INTRODUCTION mented by electroretinogramme (ERG) recordings With the exception of the extreme northern and from the surface of the eye (Kugel, 1997), demon- southern regions of the globe, phasmids can be strated that stick insects were able to make use of found in virtually any terrestrial ecosystem. How- visual signals, provided ambient light levels were ever, of the approximately 2,500 extant species, sufficiently bright for the insect’s photoreceptors to only one, the common stick insect Carausius moro- detect such signals. sus, has become a widely used laboratory species. In spite of this awareness of the stick insect’s vi- It reproduces parthenogenetically, is easy to keep, sual ability, the only existing studies to date, occu- shows interesting behavioural patterns and has pying themselves with developmental aspects of been used in studies on population ecology, genet- this insect’s photoreceptors, are those by Such ics, embryology as well as locomotion, colour (1969, 1975, 1978). Such’s research, however, cov- change, mechanoreception, and vision. ered mainly embryological and cytological aspects Adult specimens are primarily active at night and was neither concerned with developmental (Godden, 1973) and their visual capacity was origi- changes that occurred during the post-embryologi- nally considered to be poor. However, Jander and cal, nymphal life stages of the insect nor with pho- Volk-Heinrichs (1970) could show that stick in- tomechanical responses upon dark/light adaptation. sects are able to distinguish vertical patterns from Against the background of the worldwide use of C. slanted and horizontal ones and Frantsevich and morosus as a laboratory insect and the paucity of Frantsevich (1996) later discovered that a stick in- information on its postembryonic eye development sect’s preference for certain figures depended on and dark/light adaptational phenomena, a thorough whether it found itself in an upright or upside- study of these neglected aspects of stick insect

* To whom correspondence should be addressed at: E-mail: [email protected].fi

281 282 V. B. MEYER-ROCHOW and E. KESKINEN photobiology seemed overdue. overall light sensitivity of the eye, provided the photoreceptors, generally, are rather amount of the photopigment per receptor cell does flexible organs that vary widely with regard to not decline between moults. The larger size of the structure and function, reflecting species-specific eye and the greater volume of visual membrane adaptational needs as well as phylogenetic affilia- material in older individuals are likely to make tions (Land and Nilsson, 2002). Moreover, in many mechanisms capable of protecting the eye more species the eyes change in size, shape, inner organ- important. One would, therefore, expect dark/light ization, and performance as the arthropod grows adaptational phenomena to be more pronounced and matures. Further flexibility, manifesting itself and more effective in older individuals, but scien- anatomically through cell movements and appear- tific data to back up this prediction have not been ance as well as location of organelles and mem- available till now. The aim of this paper has, there- branes, assures that the eyes can cope not only with fore, been not only to document postembryonic eye the regularly recurring oscillations of day and growth in the stick insect C. morosus, but also to night, but also with sudden changes in ambient carry out a parallel investigation on the capacity light levels, temperature fluctuations and even dif- for dark/light adaptation in newly hatched and fully ferences in nutritional supplies (Meyer-Rochow, mature specimens. 1999). Eguchi and Waterman (1979) used the terms MATERIALS AND METHODS adaptive and supportive to describe the two main mechanisms involved in protecting the arthropod Individuals, representing all size and age groups eye as well as optimizing its performance. Adap- of the stick insect Carausius morosus, were housed tive mechanisms prepare the developing eye for in spacious containers in the laboratory under a particular, largely predictable, photic conditions re- light:dark regimen of 12:12 h (on: 07.00; off: lated to a species’ habitat and life-style. Except for 19.00). The temperature was maintained at a con- species living in caves, the deep-sea, and the polar stant ϩ23°C and the insects were fed a diet of regions, all experience day and night stinging nettle, dandelion, lettuce, and other sea- rhythms and their eyes, often through endoge- sonally available leaves. In order to obtain fully nously regulated processes, adjust their sensitivity dark-adapted individuals, specimens were kept in accordingly (Meyer-Rochow and Nilsson, 1999). total darkness for 24 h prior to decapitation; light- Absolute sensitivity to light is usually greater at adapted individuals faced illuminations of approxi- night, because screening pigment granules move mately 1,000 lx in intensity for 24 h prior to decapi- out of the light path, ommatidial apertures widen, tation. and amounts of visual membrane and photopig- Specimens sacrificed for transmission electron ment increase. During the day, however, reverse microscopic (TEM) observations were decapitated processes lead to sharper vision, i.e., increases in at 12.00 h and 24.00 h (referred to in the text and acuity, but at the expense of sensitivity. Sudden and figures as noon and midnight specimens, irrespec- unpredictable exposures to bright lights trigger tive of their state of adaptation). Dark adapted and faster responses than those involved in connection light adapted specimens, decapitated at both noon with the adaptive mechanisms and are primarily and midnight, are referred to as DA-noon and LA- aimed at protecting the eye against photic damage noon, and DA-midnight and LA-midnight samples, through overstimulation and dangerous forms of respectively. Each specimen was measured with a radiation, e.g., solar UV (Meyer-Rochow, 2000). ruler to its closest 1 mm and then had its severed In insects of the hemimetabolous kind, newly head immersed for 12 h at ϩ4°C in the first fixative hatched individuals usually possess tiny compound solution (i.e., 2% paraformaldehyde and 2% glu- eyes with a limited number of facets. As the indi- taraldehyde in 0.1 M phosphate buffer, buffered to a viduals grow through moulting, they add omma- pH of 7.2). Prior to postfixation in ice-cold, phos- tidia and, therefore, increase the number of pho- phate-buffered 2% osmiumtetroxide solution for toreceptive cells in the eye (Bernard, 1937). Such 30 min, the specimens were rinsed three times for changes, which often affect not only the size of the 10 min in buffer. Following postfixation, the speci- eye, but also its shape, must lead to a heightened mens were rinsed once again, but this time in dis- Eye Dynamics in the Stick Insect Carausius morosus 283 tilled water, before being passed through a graded noon and measuring 1.2, 1.5, 2.0, 2.2, 3.0, 3.5, 4.2, series of ethanol. At the completion of dehydration, 4.3, 6.0, 7.5, and 8.0 cm in body length) and an- the specimens were immersed in propylene oxide other 10 (fixed at midnight and measuring 1.3, 1.7, twice for 10 min, stepwise embedded in epon- 2.2, 3.0, 4.1, 5.5, 6.0, 7.9, and 8.0 cm in body araldite, and hardened at 60°C in an oven for two length) served as material for TEM-studies of ul- days. trastructural observations on dark/light adapta- Semithin sections for light microscopy were cut tional changes. At least 4, but most frequently 10 with a glass knife on an ultramicrotome and or more, measurements were taken on each of the stained for 10–15 s with toluidine-blue on a hot- regularly monitored anatomical features of the cen- plate. Measurements were made from digital cam- tral regions of either the left or the right eye. era images with the aid of a computer. Ultrathin The collected data were plotted against body sections were cut with a diamond knife on an ultra- lengths and R2 correlation coefficients for “Pear- microtome, picked up with carbon-coated copper son’s least-squares fit” (to test how well the meas- grids, and later stained with 2% uranyl acetate urements fitted the lines in the figures) were calcu- (20 min) and 0.4% lead citrate (5 min). The speci- lated for both noon and midnight curves. AN- mens were then observed under a transmission COVA statistics for analyses of co-variances (Rauta electron microscope at an accelerating voltage of et al., 1997) were used to calculate whether there 80 kV. Measurements were made directly on the was any statistically significant difference between images with the aid of a computer connected to the noon and midnight (*** for pϽ0.001ϭ microscope. highly significant, ** for pϽ0.01ϭsignificant and Specimens for scanning electron microscopy * for pϽ0.05ϭbarely significant). Finally, AN- (SEM) comprised animals of various sizes and COVA interaction statistics were employed to test ages, which were left to dry naturally in air. Dry whether statistically significant changes between specimens, which were noticed to have shrunk by small (body lengthsϽ2.5 cm) and large animals up to 10% when compared with live individuals, (body lengthsϾ6.0 cm) existed with regard to were placed on aluminium stubs and coated with dark/light adaptations. gold-palladium. Ommatidial counts and measure- ments on facet dimensions were made using photo- RESULTS graphic prints. For all specimens involved in the SEM-study, Morphological observations total body length (cm), head length (mm), shapes Two laterally-positioned, oval compound eyes and dimensions of the two eyes (horizontal and are present in the stick insect C. morosus right vertical principal axes of the eye in mm), average from hatching. The eyes do not display eyeshine, facet diameter (i.e., diagonal from corner to corner but at least those of the fully grown individuals ex- in mm), and total number of ommatidia per eye hibit on their surfaces a faint horizontal stripe of were recorded. Data based on measurements avail- dark, brownish coloration. Eye dimensions be- able from cross sectioned TEM-material, included tween newly hatched, first-instar stick insects, measurements on ommatidial diameter at distal measuring approx. 1.2–1.5 cm in body length, and cone level (mm), diameter of rhabdom at distal end fully grown imagines, measuring around 8 cm, vary of retina (mm), diameter of rhabdom microvilli between 0.35ϫ0.29 mm (long axisϫshort axis) and (mm), diameter of screening pigment granules 0.97ϫ0.74 mm, respectively. The general shape of (mm), and number of screening pigment granules the eye changes little as the insect matures and the inside a circum-rhabdomal annulus of half-omma- length/width ratio amounts to approx. 1.3 at all tidium width. Widths (or better thicknesses) of times. The different head/eye-length ratio of newly cone cell layers (mm) and retinal layers (mm), and hatched (3.4) and adult individuals (4.9), however, thus overall ommatidial lengths in specimens of indicates that growth follows an allometric pattern, different sizes were available from light micro- favouring head length over eye size as the insect graphs of longitudinally sectioned eyes. matures. Altogether 17 specimens were used in connec- Facet diameters increase linearly with body tion with the SEM-study. A further 11 (all fixed at length, measuring approx. 20 mm in newly hatched 284 V. B. MEYER-ROCHOW and E. KESKINEN

Table 1. Morphometric details of the eye of the stick insect C. morosus in relation to body size

Body length Head length Eye length Eye width Facet diameter Number of (mm) (mm) (mm) (mm) (mm) ommatidia

12.00 1.18 0.37 0.30 21.50 285 13.00 1.35 0.41 0.30 21.70 290 18.00 1.30 0.47 0.33 24.00 340 20.00 1.30 0.45 0.32 23.80 358 35.00 2.15 0.53 0.43 31.00 358 45.00 2.90 0.67 0.51 35.95 385 48.00 3.70 0.74 0.61 37.20 432 65.00 3.40 0.74 0.57 35.20 489 70.00 4.40 0.85 0.70 39.30 503 75.00 4.80 0.93 0.74 43.00 600 80.00 4.80 0.97 0.74 46.00 620

Anatomical features (a) General organization Each ommatidium (Fig. 3) consists of the diop- tric apparatus, i.e., cornea and cone, and the pho- toreceptive elements, i.e., retinula cells and their rhabdomeres. Throughout a stick insect’s postem- bryonic development, rhabdoms and dioptric struc- tures are not separated by a clear-zone and the eye of C. morosus, thus, conforms to the apposition compound eye type at all times. Facet widths and ommatidial lengths grow at ratios that cause in- terommatidial angles to remain more or less the same throughout a stick insect’s life. Cones are always the central products of four pigment-free cone cells. These taper proximally, Fig. 1. Facet diameters, defined as diagonal corner-to-cor- turn into cone cell roots and eventually occupy nar- ner distances in the hexagonal ommatidial lattice increase lin- row spaces between the retinula cells. Cross sec- early when plotted against body length of the stick insect. tions of the proximal ends of the cones show rhab- What the figure does not show is that facets become more and doms as ringlike, microvilli-containing structures, more regular as the insect grows. which surround the proximal cone tips. However, well over 90% of their entire length and only a lit- individuals and 45 mm in adults (Pearson’s correla- tle further proximal from the cone tips, rhabdoms tion pϽ0.001, rϭ0.696, Nϭ17: Table 1, Fig. 1). are centrally fused structures. Enveloping both The number of ommatidia per eye also increases cone and retinula cells peripherally, there are, dis- from 282 in newly hatched individuals to 600 in tally, two primary pigment cells and, more proxi- adults (Pearson’s correlation pϽ0.001, rϭ0.965, mally, an undetermined number of secondary Nϭ17: Table 1, Fig. 2). Smaller individuals have screening pigment cells. Screening pigment gran- more irregular ommatidial lattices than larger ones ules, together with other common organelles like and exhibit facets with outlines that vary from al- mitochondria, endoplasmic reticula, ribosomes, most circular to hexagonal. Even squarish and pen- vesicular material, etc., are characteristic not alone tagonal shapes are not uncommon in them. Adult of pigment cells, but of retinula cells as well. stick insects, on the other hand, possess more or (b) Dark/light adaptational changes less regularly arranged, hexagonal facets. At no de- Photomechanical events as a consequence of velopmental stage were interfacetal hairs or dark- and light-adaptations did take place and af- corneal nipples noticed. fected most noticeably the position of the screening Eye Dynamics in the Stick Insect Carausius morosus 285

Table 2. ANCOVA statistics and significance levels (*ϭweakly, **ϭmoderately, ***ϭhighly significant)

N df Adaptation Body length Interaction

Cone length 199 1 0.019* 0.000*** 0.000*** Ommatidial diameter 214 1 0.274 0.000*** 0.221 Rhabdom diameter 206 1 0.000*** 0.000*** 0.083 Retinal layer thickness 183 1 0.854 0.000*** 0.000*** Corneal thickness 196 1 0.982 0.000*** 0.971 Pigment percentage 191 1 0.000*** 0.000*** 0.002** Pigment granules 196 1 0.000*** 0.000*** 0.074 Microvillar diameter 210 1 0.573 0.677 0.054 Pigment grain diameter 201 1 0.540 0.573 0.613

Fig. 2. The total number of ommatidia grows from just about 250 in the first instar stick insects to around 600 in the adults. pigment granules around the rhabdom in the eyes of all size classes (Figs. 3, 4). But even diameters Fig. 3. Diagrammatic representation of longitudinally- and lengths of the rhabdoms themselves were influ- sectioned light-adapted (noon) eye on the left and dark- enced by environmental brightness and a host of adapted (midnight) eye on the right. Abbreviations (from top ϭ ϭ ϭ organelles, too, responded to changes in ambient to bottom): C cornea, Co cone, SPC secondary pigment cell, Ret Nϭretinula cell nucleus, Rhϭrhabdom, Pϭscreening light intensity levels with ultrastructural changes pigment, Axϭaxons. and/or translocations. However, in order to assess the effects of light and darkness quantitatively, we decided to concentrate on only those parameters cross-sectioned rhabdoms exhibited closed profiles that were readily and unambiguously measurable. and rhabdom diameters were maximal, did not This meant that we monitored changes of the diam- vary between light-adapted (noon) and dark- eters of ommatidia, pigment granules, rhabdoms, adapted (midnight) animals (ANCOVA, pϭ0.274, and microvilli, measured cone lengths and thick- dfϭ1). However, ommatidial diameters (irrespec- nesses of retinal layers and, finally, recorded the tive of adaptational state) increased linearly with distribution of screening pigment granules around body length from 16.94 mm in a 1.2 cm long, light- the rhabdom (Table 2). adapted specimen to 25.19 mm in an 8.0 cm long, Ommatidial diameters, determined at a level, dark-adapted (ANCOVA, pϽ0.001, dfϭ1; just below the proximal ends of the cones, where regression: noon pϽ0.001, rϭ0.823, Nϭ81; mid- 286 V. B. MEYER-ROCHOW and E. KESKINEN

Fig. 4. Diagrammatic representation of transversely-sec- tioned mid-rhabdom region of light-adapted (noon) eye on the left and dark-adapted (midnight) eye on the right. Abbrevia- tions: Ret Nϭretinula cell nucleus, Rhϭrhabdom, Pϭscreen- ing pigment.

Fig. 6. Rhabdom diameters (ordinate), measured from transverse sections just below the proximal tip of the cone, in light-adapted (noon) and dark-adapted (midnight) C. morosus of different body lengths (abscissa).

Fig. 5. Ommatidial diameters (ordinate), measured from transverse sections just below the proximal tip of the cone, in light-adapted (noon) and dark-adapted (midnight) C. morosus of different body lengths (abscissa). night pϽ0.001, rϭ0.887, Nϭ63) (Fig. 5). Fig. 7. Thickness of retinal layer (ordinate), measured Rhabdom diameters (Fig. 6), measured just from longitudinal sections, in light-adapted (noon) and dark- below the plane the ommatidial diameters were as- adapted (midnight) C. morosus of different body lengths (ab- scissa). Differences become obvious after the fifth moult. sessed, did vary between the light and the dark- adaptational extremes (ANCOVA, pϽ0.001, dfϭ1) from 3.68 mm in the 1.5 cm noon light-adapted ani- animals increased in body size from 1.2 cm to mal to 9.53 mm in the midnight dark-adapted 8.0 cm. Differences between light- and dark- 7.9 cm specimen. In both noon and midnight speci- adapted states, however, were only statistically sig- mens, rhabdom diameters increased linearly with nificant for specimens 6.0 cm in body length or body length (ANCOVA, pϽ0.001, dfϭ1; regres- larger (all specimens: ANCOVA, pϭ0.854, dfϭ1; sion: noon pϽ0.001, rϭ0.866, Nϭ81; midnight specimens 6 cm or larger: ANCOVA, pϭ0.032, pϽ0.001, rϭ0.895, Nϭ63). dfϭ1) and then resembled observations made ear- Retinal layers (Fig. 7) grew thicker (from 41.48 lier by Burghause (1976) on light- and dark- to 129.23 mm; regression: noon pϽ0.001, rϭ0.901, adapted ommatidia of the beetle Gyrinus natator. Nϭ81; midnight: pϽ0.001, rϭ0.926, Nϭ63) as the Cone lengths (ϭcone layer thicknesses) in- Eye Dynamics in the Stick Insect Carausius morosus 287

Fig. 8, Cone lengths (ordinate), measured from longitudi- Fig. 9. Percentage of screening pigment grains (ordinate) nal sections, in light-adapted (noon) and dark-adapted (mid- around the rhabdom of light-adapted (noon) and dark-adapted night) C. morosus of different body lengths (abscissa). (midnight) C. morosus of different body lengths (abscissa). creased from 7.09 mm in an animal of 1.5 cm body length to 25.75 mm in an 8.0 cm long specimen (ANCOVA, pϽ0.001, dfϭ1; regression: noon pϽ0.001, rϭ0.948, Nϭ81; midnight pϽ0.001, rϭ0.958, Nϭ63: Fig. 8). Noon and midnight speci- mens were statistically significantly different (AN- COVA, pϭ0.019, dfϭ1) and the interaction was strong (ANCOVA, pϽ0.001, dfϭ1). The percentage of the screening pigment gran- ules within a sleeve half the width of the ommatid- ium and surrounding the rhabdom peripherally, was calculated from counts, taken off transversely sectioned ommatidia at identical planes near the distal end of the retina just below the cones. Noon specimens were found to have a greater percentage Fig. 10. Total number of pigment granules around the of pigment granules near the rhabdom than mid- Ͻ ϭ rhabdom of transversely-sectioned light-adapted (noon) and night specimens (ANCOVA, p 0.001, df 1), but dark-adapted (midnight) C. morosus of different body lengths a smaller total number, varying between 200 to (abscissa). about 300, than the midnight specimens with ap- proximately 280–400 pigment granules (AN- COVA, pϽ0.001, dfϭ1: Fig. 9). The percentage of dfϭ1), but did vary between approx. 250 nm and pigment granules close to the rhabdom in the light- 600 nm. Pigment granule sizes did not correlate adapted specimens did not vary with body length with body length in either of the adaptational states (regression pϭ0.189, rϭ0.099, Nϭ81), but in the (ANCOVA, pϭ0.573, dfϭ1; noon regression dark-adapted specimens there was an increase of pϭ0.252, rϭ0.075, Nϭ81; midnight pϭ0.383, pigment granules near the rhabdom with increasing rϭ0.038, Nϭ63). In both adaptational stages the body length (regression pϽ0.001, rϭ0.423, Nϭ number of pigment granules per retinula cell in- 63). creased with body size (ANCOVA, pϽ0.001, The diameters of the pigment granules (ϭpig- dfϭ1; noon regression pϭ0.001, rϭ0.339, Nϭ81; ment granule size) were not found to differ in the midnight regression pϽ0.001, rϭ0.498, Nϭ63) two adaptational states (ANCOVA, pϭ0.540, (Fig. 10). 288 V. B. MEYER-ROCHOW and E. KESKINEN

thicker as the animal increased in body size (ANCOVA, pϽ0.001, dfϭ1; regression: noon pϽ0.001, rϭ0.908, Nϭ81; midnight pϽ0.001, rϭ0.928, Nϭ63).

DISCUSSION Compound eyes are present in all postembryonic developmental stages and adults of C. morosus. Through behavioural experiments it could be shown that the eyes of the adult individuals (Jander and Volk-Heinrichs, 1970), and at least the last two larval instars as well (Frantsevich and Frantsevich, 1996), were functional. There is no reason to as- sume that any of the earlier instars would not also Fig. 11. Diameters of rhabdom microvilli (ordinate) in use their eyes. Although, as the insect grows, there light-adapted (noon) and dark-adapted (midnight) C. morosus is little change in the shape of the eye, there are of different body lengths (abscissa). nearly threefold increases affecting number of facets, ommatidial diameter and corneal thickness. Rhabdom diameters and the thickness of the retinal layer also increase, but only by a factor of two. These growth-related morphological changes will be discussed first and an assessment of the func- tional consequences of the observed dark/light adaptational changes will then follow. The somewhat slower rate of growth of the eye in relation to body length is not unusual and has been documented for several species of insects (Bernard, 1937). However, the greater number of facets coupled with larger ommatidial diameters gives the adult an approximately 10 times more ex- tensive surface area. This, even if the amount of photopigment per ommatidium were to remain constant as the animal grows, would provide the Fig. 12: Corneal thicknesses (ordinate) in light-adapted adults with a considerably increased sensitivity to (noon) and dark-adapted (midnight) C. morosus in relation to body length. light. Furthermore, given that the interommatidial angle changes little between nymphs and adults, the larger facet diameters of the adults would also The diameters of the rhabdom microvilli (Fig. improve resolution, for the greater the aperture of 11) were not only found to maintain their value of the ommatidium, the smaller the so-called blur cir- ϭ approx. 0.45 nm throughout the entire postembry- cle (defined as Da l/Dl, where l is wavelength onic growth of the stick insect, but also to remain and Dl, in radians, is the diameter of the lens, in unaffected by dark/light adaptations (ANCOVA, this case the ommatidium: Horridge, 1978). The F- interaction pϭ0.054, body length pϭ0.677; adap- number of an optical system is an indicator of the tation pϭ0.573, dfϭ1; noon regression pϭ0.063, relationship between the two most important func- rϭ0.192, Nϭ81; midnight pϭ0.257, rϭ0.084, tional parameters of any organ of vision, namely Nϭ63) (Fig. 11). sensitivity and resolution. If the F-number, defined

The thickness of the cornea (Fig. 12) did not in insects as the quotient f/Dl (where f is focal dis- change between noon and midnight specimens tance and Dl is the diameter of the ommatidial (ANCOVA, pϭ0.982, dfϭ1), but grew linearly lens) decreases, sensitivity increases (Land, 1981). Eye Dynamics in the Stick Insect Carausius morosus 289

Since resolution in a compound eye is defined as however, the diameters of the microvilli change ϭ R r/Dl (where r is the local optical radius of the only insignificantly. Presuming that photopigment eye and Dl is the functional diameter of the corre- density remains the same in nymphs and adults ϭ sponding ommatidium) and DF Dl·rradians, then (and there is no evidence to suggest that it does the product Dl·DF, called the eye parameter ‘p’, change), then the increase in rhabdom bulk more becomes a measure of how much resolution of than likely should be capable of compensating for each lens is sacrificed to sensitivity by increases in the losses in photon flux that might be caused by field and receptor sizes if the eye is to be able to the thicker corneae and the self-screening proper- see objects at low light intensities. With a DF of ties of the elongated rhabdoms in the adults (Gold- 6°, the eyes of the smallest nymphs would operate smith, 1978). Indirect evidence that increases in with a p-value of 2.1, but the adults, having a DF rhabdom diameters and rhabdom lengths improve of 5.5° and facets of at least 40 mm in diameter, sensitivity, stem from the fact that the widest and would have an eye parameter of 3.8. This demon- largest rhabdoms are present in stick insects fixed strates that adults would be able to perceive a simi- at midnight, when ambient light levels reach a min- lar amount of detail at considerably dimmer ambi- imum. Increases in rhabdom diameters, moreover, ent light levels than younger and smaller individu- correlate well with increased sensitivity in numer- als. ous species of arthropods, albeit at the expense of But how much is there to see for first instar and acuity (Warrant and McIntyre, 1993). However, the adult C. morosus? To answer this question we need greater regularity with which the facets of the to know the receptor acceptance angle Dr and that adults (compared with those of the early instars) depends significantly on the angle subtended by the cover the surface of the eye, makes amends to any distal rhabdom end in the outside world d/fradians, possible loss in acuity. Deviations from regularity, where d is rhabdom diameter and f is focal distance even if the total facet surface area were to remain or more precisely posterior nodal point of the lens unaltered, are clearly disadvantageous in modula- (Horridge, 1978). Since the F-number is defined as tion transfer functions (French et al., 1977) and fre- f/D, we can simplify the equation for the receptor quently affect the organization below the aberrantly ϭ acceptance angle to Dr d/DFradians, where d is dis- shaped facets (Keskinen et al., 2002). tal rhabdom diameter, D is facet diameter, and F is The larger size of the eye and consequently the F-number. If we assume the F-number of the stick greater absolute volume of visual membrane mate- insect eye to be typical of that found in other appo- rial in the fully grown insect is likely to make sition eyes, namely 2, the receptor acceptance mechanisms capable of protecting the eye more angle becomes a function of facet diameter and important. The reasons are twofold. Firstly, bigger rhabdom diameter. The two variables have been and more facets mean that a greater amount of po- measured by us for both first instar and mature in- tentially damaging radiation can enter the eye dividuals under day as well as night conditions. (Meyer-Rochow et al., 2002), since the dioptric Consequently, respective receptor acceptance an- structures of insect photoreceptors are poor UV-fil- gles for the small eyes of the young and the large ters (Meyer-Rochow, 1975). Secondly, the closer a eyes of the old stick insects are 5.3° and 4.7° dur- growing individual approaches its final moult, the ing the day, and 8° and 7.3° at night. These results smaller is its chance of being able to repair any clearly demonstrate that in addition to the theoreti- damage in a future moult. One would, therefore, cally possible great gain in sensitivity, small im- expect dark/light adaptational phenomena to be provements in resolution can also accompany eye more pronounced and more effective in mature in- growth in C. morosus. dividuals. To some extent our data support this no- Are there any reasons to doubt that the theoreti- tion: total amounts of screening pigment granules cally possible gains in sensitivity and acuity cannot do increase as the eye increases in size and actually be realized? We have shown that in paral- dark/light adaptational changes in the width of the lel with the bodily growth of C. morosus, facets not retinal layer become significantly more pronounced only build up numbers, but grow in size, that reti- in mature individuals (cf. also Insausti and Lazzari, nal layers become wider, rhabdoms lengthen, and 2000). The reason why transversely-cut rhabdoms rhabdom diameters increase; at the same time, of dark-adapted stick insects seemingly possess 290 V. B. MEYER-ROCHOW and E. KESKINEN more screening pigment grains in their retinula are less sensitive to light than those of the adults. cells (although further away from the rhabdom’s In fact, the small eyes of first instars and their re- edge than under conditions of light adaptation) is, sultant low sensitivity to light may help the tiny, because at light adaptation most of the pigment mi- newly emerged individuals to escape from the dark grates from the middle of the retinula cells to the leaf letter and to move from the location of the extreme distal ends, thereby forming a narrow eggs into the illuminated zone of edible foliage. sleeve around the aperture (i.e., the proximal cone Adults, being better able than the younger stages to tip) of the ommatidium. Although changes affect- adjust their vision to the dim ambient light condi- ing rhabdom diameters are equally well developed tions prevailing during the time of night and at in first instars and adults, the trend of somewhat dusk, are almost entirely nocturnal (Godden, narrower rhabdom microvilli to be present in dark- 1973). Early stages, however, will move freely adapted adults would favour heightened sensitivity, around and feed during the day, thereby expanding as it would result in increased membrane material their range. Since adults of the water stick insect and, thus, allow more photopigment to be accom- Ranatra measure approximately 3.5 cm in length modated in the eye at night. and exhibit visual reactions to objects as far away Are there behavioural tests or observations to as 46 mm (Cloarec, 1984b), we can probably as- prove some of the predictions made on the basis of sume that 1.8 cm long first instar C. morosus and our morphological findings? Stick insects have to 8 cm long adults can detect objects at least 2 and be able to see, approach, and reach for thin 10 cm away, respectively. Whether young and older branches and leaves. They are known to respond stick insects differ with regard to flicker fusion fre- differently to horizontal, vertical, or inclined quencies is unknown, but reported flicker fusion stripes as well as plain or broken-up patterns (Jan- frequencies in the adults of around 15/s (Kaestner, der and Volk-Heinrichs, 1970). They are also 1972) are very low when compared with those of known to exhibit visual resolutions that are much other insects: obviously, for a relatively slow and below the theoretical limits predicted by the in- non-flying insect like C. morosus, selective pres- terommatidial angle and receptor spacing (Jander sure to evolve high temporal resolution was clearly and Volk-Heinrichs, 1970). This allies them with less effective than the need to adjust spatial resolu- other insects, like Velia (Meyer, 1974) and Ranatra tion to low light levels. (Cloarec, 1984a), and is probably linked to a fre- ACKNOWLEDGEMENTS quently observed gentle swinging of the body from side to side, permitting the insect to make use of EK wishes to acknowledge the support received towards the parallaxis (Horridge, 1977). Ranatra, the water completion of this project by the “Jenny and Antti Wihuri stick insect, does not approach and climb branches, Fund” (Jenny ja Antti Wihurin rahasto) and the Finnish Konkordia Fund (Suomalainen Konkordia-liitto). We are in- but it sees and seizes prey within its visual field. debted to Mr. Pekka Tokola (Oulun Normaalikoulu; Finland) Both Ranatra and C. morosus have a common vi- for providing us with stick insects originally from a culture sual problem in spite of the different life styles maintained by Prof. F. Weber (University of Münster; Ger- they lead and their visual systems are thus compa- many). rable. Surprisingly, the rate of increase in facet REFERENCES numbers between first instars and adults of Ranatra and the increases in facet diameter are quite similar Bernard, F. (1937) Recherches sur la morphogénèse des yeux to those seen in the stick insect. At the same time it composés d’arthropodes: développement, croissance, re- must be said, however, that interommatidial angles duction. Suppl. Bull. Biol. Belg. 23: 1–166. Burghause, F. (1976) Adaptationserscheinungen in den decrease in the growing Ranatra eye, but stay more Komplexaugen von Gyrinus natator L. (Coleoptera: or less the same in C. morosus. This means that in Gyrinidae). Int. J. Morphol. Embryol. 5: 335–348. the stick insect improvements in visual acuity stem Cloarec, A. (1984a) Development of the compound eyes of mostly from an increase in facet width and not a the water stick insect, Ranatra linearis. Physiol. Ento- decrease in interommatidial angle, a pattern that mol. 9: 253–262. Cloarec, A. (1984b) Mecanismes intervenant dans l’estima- resembles that seen in dragonflies (Sherk, 1978). tion des distances predatuerproie chez Ranatra. Behav. The behaviour of early instar stick insects is in Proc. 9: 123–133. full agreement with the conclusion that their eyes Eguchi, E. and T. H. Waterman (1979) Longterm dark in- Eye Dynamics in the Stick Insect Carausius morosus 291

duced fine structural changes in crayfish photoreceptor isierung des optischen Abtastrasters beim membrane. J. Comp. Physiol. 131: 191–203. Bachwasserläufer (Velia caprai). J. Comp. Physiol. 92: Frantsevich, L. and L. Frantsevich (1996) Space constancy 85–103. in form perception by the stick insect. Naturwis- Meyer-Rochow, V. B. (1975) The dioptric system in beetle senschaften 83: 323–324. compound eyes. In The Compound Eye and Vision in In- French, A. S., A. W. Snyder and D. G. Stavenga (1977) sects (G. A. Horridge ed.). Clarendon Press, Oxford, Image degradation by an irregular retinal mosaic. Biol. pp. 299–313. Cybernetics 27: 229–233. Meyer-Rochow, V. B. (1999) Compound eye: Circadian Godden, D. H. (1973) A re-examination of circadian rhyth- rhythmicity, illumination, and obscurity (Chapter III-2). micity in Carausius morosus. J. Insect Physiol. 19: In Atlas of Arthropod Sensory Receptors (E. Eguchi and 1377–1386. Y. Tominaga eds.). Springer-Verlag, Tokyo, pp. 97–124. Goldsmith, T. H. (1978) The spectral absorption of crayfish Meyer-Rochow, V. B. (2000) Risks, especially for the eye, rhabdoms: pigment, photoproduct, and pH-sensitivity. emanating from the rise of solar UV-radiation in the Arc- Vision Res. 18: 463–473. tic and Antarctic regions. Int. J. Circumpolar Health Horridge, G. A. (1977) Insects which turn and look. En- 59: 38–51. deavour 1: 7–17. Meyer-Rochow, V. B., T. Kashiwagi and E. Eguchi (2002) Horridge, G. A. (1978) A different kind of vision: the com- Selective photoreceptor damage in four species of insects pound eye (Chapter 1). In Handbook of Perception. Vol. 8 induced by experimental exposures to UV-radiation. (E. C. Carterette and M. P. Friedman eds.). Academic Micron 33: 23–31. Press, New York, pp. 3–82. Meyer-Rochow, V. B. and H. L. Nilsson (1999) Compound Insausti, T. C. and C. Lazzari (2000) An ocellar pupil that eyes in polar regions, caves, and the deep-sea (Chapter does not change with light intensity, but with the insect’s III-3). In Atlas of Arthropod Sensory Receptors (E. age in Triatoma infestans. Mem. Inst. Oswaldo Cruz 95: Eguchi and Y. Tominaga eds.). Springer-Verlag, Tokyo, 743–746. pp. 125–142. Jander, R. and I. Volk-Heinrichs (1970) Das strauch-spezifis- Rauta, E., H. Rita and J. Kouki (1997) Biometria. che visuelle Perceptor-System der Stabheuschrecke (Ca- Yliopistopaino, Helsinki. rausius morosus). Z. Vergl. Physiol. 70: 425–447. Sherk, T. E. (1978) Development of the compound eyes of Kaestner, A. (1972) Lehrbuch der speziellen Zoologie, Band dragonflies (Odonata), II. Development of the larval com- I: Wirbellose 3.Teil. Gustav Fischer Verlag, Stuttgart. pound eyes. J. Exp. Zool. 203: 47–60. 272 pp. Such, J. (1969) Observations sur l’infrastructure des cellules Keskinen, E., Y. Takaku, V. B. Meyer-Rochow and T. cristalliennes de l’oeil du phasme Carausius morosus Br.: Hariyama (2002) Microanatomical characteristics of ces cellules sont-elles de simples elements “dioptriques”? marginal ommatidia in three size-classes of the semi-ter- C. R. Acad. Sci. Paris 269: 356–359. restrial isopod Ligia exotica (Crustacea; Isopoda). Bio- Such, J. (1975) Analyse ultrastructurale de la morphogénèse cell 26: 357–367. ommatidienne au cours du développement embryonnaire Kugel, M. (1977) The time course of the electroretinogram de l’oeil compose chez le phasme Carausius morosus of compound eyes in insects and its dependence on spe- Br. C. R. Acad. Sci. Paris 281: 67–70. cial recording conditions. J. Exp. Biol. 71: 1–6. Such, J. (1978) Embryologie ultrastructurale de l’ommatid- Land, M. F. (1981) Optics and vision in invertebrates. In ium chez le phasme Carausius morosus Br. (Phasmida: Handbook of Sensory Physiology. Vol. VII/6B (H. Lonchodidae): morphogénèse et cytodifférenciation. Autrum ed.). Springer Verlag, Berlin, pp. 471–592. Int. J. Insect Morphol. Embryol. 7: 165–183. Land, M. F. and D.-E. Nilsson (2002) Animal Eyes. Oxford Warrant, E. J. and P. D. McIntyre (1993) Arthropod eye de- University Press, New York. 221 pp. sign and the physical limits to spatial resolving power. Meyer, H. W. (1974) Geometrische und funktionelle Spezial- Progr. Neurobiol. 40: 413–461.