Absolute and Spectral Sensitivities in Dark- and Light-Adapted Pagothenia Borchgrevinki, an Antarctic Nototheniid Fish

Physiology & Behavior, Vol. 61, No. 2, pp. 159–163, 1997 Copyright q 1997 Elsevier Science Inc. Printed in the USA. All rights reserved 0031-9384/97 $17.00 / .00

PII S0031-9384(96)00354-X

Absolute and Spectral Sensitivities in Dark- and Light-Adapted Pagothenia borchgrevinki, an Antarctic Nototheniid Fish YUKITOMO MORITA,* V. BENNO MEYER-ROCHOW† 1 AND K. UCHIDA* *First Department of Physiology, Hamamatsu University School of Medicine, Hamamatsu 431-31 Japan and †Department of Biology, Section Animal Physiology, University of Oulu, SF-90571 Oulu, Linnanmaa, Finland Received 13 February 1995; Accepted 19 June 1996 MORITA, Y., V. B. MEYER-ROCHOW AND K. UCHIDA. Absolute and spectral sensitivities in dark- and light-adapted Pagothenia borchgrevinki, an Antarctic nototheniid fish. PHYSIOL BEHAV 61(2) 159–163, 1997.—Functional properties of the retina of Pagothenia borchgrevinki, an Antarctic nototheniid fish that lives beneath the 2.5–3 m thick sea-ice in water of 01.87C temperature, were analyzed electrophysiologically at Scott Base (777 50 * S; 1667 45 * E). The waveform of the ERG was monophasic in the dark-adapted state and showed an off-response of opposite polarity in the light-adapted condition. Responses of the light-adapted retina were smaller than those of the dark-adapted eye, although both photopic and scotopic components were observed. Spectral sensitivity measured by monochromatic photostimulation at 14 different wavelengths across the 400–700-nm range showed a single maximum at 490 nm. The spectral sensitivity curve is consistent with a rhodopsin photopigment. The darkadapted retina exhibited a photon flux density threshold of approximately 2 1 10 9 photons cm02 s 01 ) when monochromatic flashes of 500 nm wavelength and 250 ms duration were used. When the stimulus consisted of 1 s white light, a minimum energy flux density of approx. 2 1 10 04 mW/cm2 was necessary to elicit a detectable response. It was concluded that the visual system of P. borchgrevinki was in tune with the dominant downwelling spectral irradiance and that, due to retinal thermal noise reduction in the cold environment, no great need for particular anatomical adaptations to further enhance sensitivity existed. Copyright r 1997 Elsevier Science Inc. Spectral and absolute sensitivity

Electrophysiology

Retina

Fish eye

Antarctica

Pagothenia borchgrevinki

common nototheniid fish in the McMurdo Sound / Antarctica ) it was concluded that visually mediated feeding would be limited to the top 20 – 40 m of the water column ‘‘under average spring conditions of snow and ice cover’’ ( 15 ) , and that the lateral line organ replaced the visual sense at low ambient light levels ( 13 ) . We have recently repeated and expanded some of the physiological measurements of spectral and absolute sensitivities in P. borchgrevinki and conclude that the eyes of this fish can function very well at light levels earlier deemed too dim for vision ( 15 ) .

BECAUSE of the extremely difficult working conditions under which research in Antarctica is carried out, very few direct observations on vision in Antarctic fishes exist. This is somewhat disappointing because 1. the Antarctic realm makes up one of the major biota on earth, and 2. the marine environment of high southern latitudes, even when compared with that of the Arctic ( 5 ) , is unique because of constantly low temperatures, low ambient light levels, and stable salinity conditions ( 6 ) . Light transmission in the Antarctic sea is influenced by a variety of factors, including sun angle, snow cover, ice thickness, ice impurities, epontic phytoplankton, etc. However, even during the Antarctic summer, with continuous daylight for several months, light intensities measured under 3 m of solid sea-ice covered by 5 cm snow gave only a noon average light transmission of 0.25% ( 7 ) . Pankhurst and Montgomery ( 14 ) plotted depth light-intensity profiles at different times of day and found highest light intensities at 1500 h with measurable light penetration ( equivalent to 1.2 1 10 11 photons cm02 s 01 ) down to 40 m depth. On the basis of electrophysiologically and behaviorally determined visual thresholds in Pagothenia borchgrevinki ( a

METHOD

Specimens of the Antarctic nototheniid fish Pagothenia borchgrevinki (Boulenger) were caught by hook and line from an approx. 1-m wide hole through the 2.5–3 m thick sea-ice 1.5 km to the south of Scott Base (777 50 * S; 1667 45 * E). Fish (ca. 20 cm in length) were carried to the laboratory in a black, lighttight, 20 l plastic container filled with cold seawater. Following dark-adaptation of at least 3–5 h in water of 07C, the eye was excised under dim red light at room temperature (10–137C) from

1

To whom requests for reprints should be addressed. E-mail: [email protected]

159 / eh08 2475 Mp 159 Wednesday Jan 08 04:12 PM EL–PB (v. 61, No. 2) 2475

160

MORITA, MEYER-ROCHOW AND UCHIDA for photometric calibrations. A photon flux density of 6.24 1 10 13 photons cm02 s 01 ( corresponding to an energy flux density of 24.8 mW / cm2 ) was available for the 500-nm test light, and the maximum energy flux density for the white test light without attenuation by ND filters was calculated to amount to approx. 2 1 10 3 mW / cm2 . The potential recorded was preamplified by a specially designed differential preamplifier with low noise level of 3 mV and bandpass filter width of 0.15–100 Hz. A digital storage oscilloscope with functions for averaging responses (VP-5730A, 50 MHz, National Co. Ltd., Japan) was used to observe and to measure the photic responses. A just-measurable deflection of the ERG-trace was taken as an indication of the threshold. Permanent photographic records of the potentials were obtained by polaroid camera. ERGs of light-adapted eye-cup preparations were recorded under sustained illumination of either 12 or 190 lx to which the stimulating flash was added. Each value, whether from dark- or light-adapted preparations, was based on at least 3 individual recordings. RESULTS

FIG. 1. Electroretinogram of P. borchgrevinki eye-cup preparation in dark-adapted (A) and light-adapted (B) states. Test light in (A): 500 nm, 250 ms (12.4 mW/cm2 Å 3.12 1 10 13 photons cm02 s 01 , corresponding to 1.6 1 10 4 times threshold). Test light in (B): white light, 250 ms, 1900 lx (equivalent to approx. 2 1 10 3 mW/cm2 ) and applied under sustained illumination of 12 lx white light (equivalent to approx. 10 mW/cm2 and corresponding to 6.3 1 10 4 times threshold level). Calibration: 10 mV and 200 ms in grid for both records. Response range: 0.15–100 Hz. Negative Å down.

fish that had been decapitated with a sharp knife and left a further 60 min to dark-adapt. The eye cup preparation was installed on Ringer-soaked filter paper in the recording chamber. The electroretinogram (ERG) was recorded with glass electrodes of 10 mm tip diameter, filled with Ringer-Agar and a thin Ag-AgCl wire for contact with the recording cable. The small Ag-AgCl plate beneath the eye cup served as the indifferent electrode. A lightproof Faraday cage was used to reduce electrical noise. The photoelectric artifact was checked and it was confirmed that no potential appeared even under the most intense illumination. A light-proof ventilator was used to keep the temperature inside the dark Faraday cage as cold as the laboratory temperature (10– 137C). This temperature is not excessive, considering that propagation of action potentials is blocked at about 287C in Antarctic fishes (8) and endplate potential amplitudes in extraocular muscle of P. borchgrevinki remain unchanged between 02 and 0157C (16). Most of the equipment, such as the portable monochromatic stimulator with 16 interference filters of half-band widths of 15 nm (Nihon-shinku-kohgaku Co. Ltd., Japan), was specially designed to work in Antarctica. A 15-V, 150-W ellipsoid halogen reflector bulb (Schott KL-1500 ) served as the light source. The active range of the stimulator was 400–700 nm. Neutral density filters (ND-wedge and ND-filters: Chuo-seiki Co. Ltd., Japan) were applied to attenuate the photic energy. An electronic stimulator triggered an electromagnetic shutter and the sweep of the cathode ray oscilloscope. A radiometer ( RM101: Tokyo-kohdensi-kogyo Co. Ltd., Tokyo, Japan) , a vacuum thermopile ( type G: Pyro-Werk, Hannover, Germany ) and an ICE spot-photometer were used

The signal / noise ratios of typical ERGs from dark- and light-adapted P. borchgrevinki eyes are shown in Fig. 1A, B. The fully dark-adapted retina responded to light with a negative deflection, resembling the combined receptor potential and glial response ( Fig. 1A ) ; b-waves were not observed. A test light of 500 nm, 3 1 10 13 photons cm02 s 01 evoked a large response with a steep rising phase. The light-adapted ERG had

FIG. 2. Recordings from dark- (A) and light-adapted (B) states. (A): Increasing light intensities evoke larger responses with steeper rising phases and shorter latencies. The 250 ms test light of 500 nm was increased as follows: 3.1 1 10 10 ; 6.2 1 10 10 ; 5 1 10 11 ; 4 1 10 12 ; 3 1 10 13 photons cm02 s 01 . (B): Light-adapted responses from 2 preparations. Off responses are separated by longer illumination. The largest response of 2(A) and trace 1 of 2(B) correspond to the recordings shown in Figs. 1A and B, respectively.

/ eh08 2475 Mp 160 Wednesday Jan 08 04:12 PM EL–PB (v. 61, No. 2) 2475

VISUAL PROPERTIES OF ANTARCTIC FISH EYE

161 DISCUSSION

FIG. 3. Intensity/response relationships in dark- and light-adapted states. Filled symbols represent dark- and open symbols light-adapted preparations. Test lights of 500 nm and 250 ms duration were used in the case of the dark-adapted retina; white light was used for the light-adapted eye. The conditioning light for light adaptation was white light of 12 lx (see Fig. 1B). The value of 0.0 on the abscissa corresponds to 24.8 mW/cm2 ( Å 6.24 1 10 13 photons cm02 s 01 ) in the case of the 500 nm test light and 1900 lx (approx. 2 1 10 3 mW/cm2 ) in the case of the white test light.

a slower rising phase, even if the white test light was as bright as 2 1 10 3 mW / cm2 . A large ’off ’ response was characteristic of the light-adapted ERG. In general, stimulation by intense light was necessary to obtain a response following light adaptation. The effect of increasing light intensity is shown in Fig. 2A. As the amplitude of the response increased, the latency became shorter and the rise phase steeper. In Fig. 2B, two stimuli differing in length ( L1 and L2 ) are applied to show the separation of the ’off ’ responses and that the latter are not rebounds of the negative ’on’ responses. Response amplitudes increased significantly in the darkadapted retina (indicative of a broader dynamic range) but, in the light-adapted retina, the slope of the response increase was more gentle. The intensity/response relationships of some typical examples are shown in Fig. 3. The spectral sensitivity curve of the ERG, based on responses to 14 different wavelengths, is shown in Fig. 4. The log relative sensitivity of the dark-adapted retina was plotted following measurements of the energy necessary to give a constant amplitude from the linear part of the intensity / response curve, as shown in Fig. 3. Following comparison with nomograms ( 3 ) , we conclude that the recorded spectral sensitivity curve is characteristic of a rhodopsin with peak sensitivity at 490 nm. In 6 of 8 cases, ERG-thresholds, based on stimulation with 500-nm monochromatic light, were 2 1 10 9 photons cm 02 s 01 ; in the remaining 2, they were 6 1 10 9 photons cm 02 s 01 . When white light stimuli were used, in 3 of 3 tests, thresholds were reached at energy flux densities of approx. 2 1 10 04 mW / cm 2 .

The existence of both rod and cone photoreceptor cell types in the retina of P. borchgrevinki was first reported by MeyerRochow and Klyne ( 10 ) . We are aware that different response waveforms need not conclusively demonstrate that different receptor types are active ( dark- and light-adapted rods can have very different response kinetics ) , but our adapting background lights were so strong that the rods were expected to have become saturated. We are, therefore, inclined to believe that the light-adapted ERG was cone-driven. Adaptational changes were very slow and, thus, may not be of any great importance to the animal during the Antarctic summer. The recorded spectral sensitivity curves of the darkadapted P. borchgrevinki eye are similar to those obtained earlier ( 14 ) and, regrettably, also do not cover the ultraviolet range. UV-levels under 3 m of solid ice, however, are expected to be very low and of considerably less importance than the spectral range examined. Given that the earlier study was based on only 6 gelatin color filters to cover the entire visual spectrum, its conclusion of 482 nm as peak sensitivity and ours of 490 nm are in surprisingly good agreement. Bluegreen light peak sensitivities have been reported from the eyes of other marine Antarctic organisms [ fish Trematomus spp. ( 14 ) ; crustacean Orchomene plebs ( 9 ) ] and it seems that

FIG. 4. Spectral sensitivity curve based on averages of ERG-recordings from 6 different fish retinas. The log of the relative sensitivity (with SD) of the dark-adapted retina was plotted following measurements of the energy necessary to give a constant amplitude of ERG. Based on Dartnall’s (3) nomogram, the curve is characteristic of a rhodopsin photopigment with peak sensitivity of 490 nm.

/ eh08 2475 Mp 161 Wednesday Jan 08 04:12 PM EL–PB (v. 61, No. 2) 2475

162

MORITA, MEYER-ROCHOW AND UCHIDA

spectral sensitivity is in tune with the dominant downwelling spectral irradiance, which peaks at 490 – 550 nm under hard ice with minimal platelet formation ( 17 ) . Where our results differ considerably from the earlier study ( 14 ) is in relation to threshold sensitivities. Based on eliciting a feeding response with white light in the laboratory, Pankhurst and Montgomery (14) gave a figure of 3 1 10 11 photons/cm02 s 01 as threshold intensity (recalculated by us from their original figure of 5 1 10 03 mE m02 s 01 ). Measurements of visual thresholds, however, are not easy and depend considerably on the method employed to detect the response. Feeding, for instance, before leading to pursuit and uptake of the prey, could be preceded by eye, rather than body, movements, which are easily overlooked. Using a low-noise preamplifier with particularly good signalto-noise ratio, we employed as a criterion the smallest measurable potential change of a detectable signal and found that, for bluegreen light of 500 nm wavelength, approx. 2 1 10 9 photons m02 s 01 constituted the threshold light intensity. Assuming that vertical light transmission in Antarctic water is not markedly different from that of other water bodies, we conclude that vison should be possible for P. borchgrevinki beyond the 20–40 m and 30–60 m depth that was predicted as the limit by Pankhurst and Montgomery (14) for photopic and scotopic vision, respectively. The conclusion (14) was reached on the basis of depth/intensity profiles and a measured photon flux density of 1.2 1 10 11 photons cm02 s 01 at 40 m depth (recalculated by us from the original figure of 0.002 mE m02 s 01 ). Despite our ERG-thresholds being approx. 2 log units below those of Pankhurst and Montgomery (13), we expect them to be still far above the real functional visual threshold of the P. borchgrevinki eye. First, they only reflect the smallest signal we could discriminate from baseline noise and are in that sense ‘arbitrary.’ Second, the capacity of the retina and the visual system of an animal to sum and distinguish small signals depends to a large extent on neural factors that we do not yet know. Third, according to Donner (personal communication), in the frog the absolute visual threshold would typically correspond to an ERG mass receptor response of ca. 1% of the maximal response. If this was applied to P. borchgrevinki as well, it would result in a 5 times lower threshold than what even we had measured and, using the depth/profile formula given (14), would extend vision into depths 2 to 3 times greater than was thought possible previously. The conclusion that the fish depend on ‘‘nonvisual systems over a substantial part of their distribution range’’ (14) may, therefore, have been premature. Compared to the previously reported visual thresholds of P. borchgrevinki ( 14 ) , our results may at first seem surprising, given that, anatomically, the eye ( and in particular the retina ) of P. borchgrevinki does not exhibit any obvious adaptations to a dimly lit environment to enhance sensitivity ( 4,11 ) . However, a fish that lives in water of constantly subzero tempera-

tures [ yearly average 01.817C ( 7 ) ] may not need any special anatomical adaptations because low environmental temperature reduces thermal noise in the retina ( 1 ) . Even the terrestrial, nocturnally feeding New Zealand tuatara ( Sphenodon punctatus ) still responded to prey at a light level of 6 mlux ( 12 ) . Aho et al. ( 2 ) commented on reported frog ‘‘visual thresholds’’ of 10 mlux, and the visual threshold of P. borchgrevinki to white light ( here expressed as 2 mlux to facilitate comparisons ) does not seem out of order or even especially high. Irradiance in the environment, of course, cannot be equated with that on the retina, but data on the transmission ratio of dioptric structures, such as cornea, lens, and vitreous humor, are not available. Moreover, these structures are so clear that, over the short distance the light has to travel in them, no signficant attenuation affecting our conclusions is expected to take place. What, then, could have caused earlier threshold values ( 14 ) to be so different from ours? Most important is, first of all, that sufficiently dark-adapted fish are being used. To achieve complete dark adaptation in the frigid Antarctic water may take a long time [ several days in the case of the Antarctic crustacean Glyptonotus antarcticus ( 10 ) ] and newly captured fish should be shielded from the light as soon as they are pulled through the ice-hole. Second, our equipment contained an especially good preamplifier with high signal / noise ratio that permitted us to detect small responses otherwise buried in noise. Third, the detection of the smallest deflection was facilitated by being able to average up to 4 measurements at any one time of a recording. Provided the correct precautions are taken and the most suitable equipment is selected, we predict that even lower visual thresholds are likely to be present in Trematomus bernacchii and Dissostichus mawsoni, fishes that inhabit greater depths of McMurdo Sound and possess a larger number of rods in the retina than P. borchgrevinki (11). ACKNOWLEDGEMENTS

The authors wish to express their appreciation to the staff of the D. S. I. R. ( Antarctic Division ) for the assistance and hospitality received at Scott Base. The work was carried out during the expedition KO22 of the Waikato University Antarctic Research Unit ( WARP ) and funded by grants from the U. G. C. to V. B. M-R. and grant-in-aid for the International Joint Research on Pineal Photoreception from the Ministry of Education, Science, and Culture of Japan ( no. 63044059 ) to Y. M. Finally, our sincere thanks go to Prof. N. Maruyama and Dr. Y. Taga of the Brain Institute ( Niigata University, Japan ) for preparation of the preamplifier and recorder used in this investigation. V. B. M-R. is indebted to Prof. M. Ja¨rvilehto ( Oulu University, Finland ) , in whose lab the manuscript was completed during tenure of a Finnish Academy Visiting Professorship, and to Dr. K. Donner ( Helsinki University, Finland ) for invaluable comments to improve the text.

REFERENCES 1. Aho, A. C.; Donner, K-O.; Reuter, T. Retinal origin of the temperature effect on absolute visual sensitivity in frogs. J. Physiol. London 463:501–523; 1993. 2. Aho, A. C.; Donner, K-O.; Hyden, C.; Larsen, L. O.; Reuter, T. Low retinal noise in animals with low body temperature allows high visual sensitivity. Nature 334:348–350; 1988. 3. Dartnall, H. J. A. The interpretation of spectral sensitivity curves. Br. Med. Bull. 9:24–30; 1953. 4. Eastman, J. T. Ocular morphology in Antarctic notothenioid fishes. J. Morphology 196:283–306; 1988.

5. Hedgpeth, J. W. Perspectives in benthic ecology in Antarctica. In: Quam, L. O., ed. Research in the Antarctic. Washington: AAA Science; 1971:93–136. 6. Jacobs, S. S.; Amos, A. F.; Bruchhausen, P. M. Ross Sea oceanography and Antarctic bottom water formation. Deep Sea Res. 17:935–962; 1970. 7. Littlepage, J. C. Oceanographic investigations in the McMurdo Sound, Antarctica. Biology of Antarctic Seas 2. Antarctic Res. Ser. 5:1–37; 1965. 8. Macdonald, J. A. Temperature compensation in the peripheral nervous system: Antarctic vs. temperate poikilotherm animals. J. Comp. Physiol. 142:411–418; 1981.

/ eh08 2475 Mp 162 Wednesday Jan 08 04:12 PM EL–PB (v. 61, No. 2) 2475

VISUAL PROPERTIES OF ANTARCTIC FISH EYE 9. Meyer-Rochow, V. B. Eye colour and spectral sensitivity in the Antarctic amphipod Orchomene plebs. NZ Antarctic Rec. 3:25–29; 1980. 10. Meyer-Rochow, V. B. The divided eye of the isopod Glyptonotus antarcticus: effects of unilateral dark adaptation and temperature elevation. Proc. Roy. Soc. London B215:433–450; 1982. 11. Meyer-Rochow, V. B.; Klyne, M. A. Retinal organization of the eyes of three nototheniid fishes from the Ross Sea (Antarctica). Gegenbaurs Morphol. Jahrbuch 128:762–777; 1982. 12. Meyer-Rochow, V. B.; Teh, K. L. Visual predation by tuatara (Sphenodon punctatus) on the beach beetle (Chaerodes trachyscelides) as a selective force in the production of distinct colour morphs. Tuatara 31:1–9; 1991. 13. Montgomery, J. C.; Macdonald, J. A. Sensory tuning of lateral line receptors in Antarctic fish to the movements of planktonic prey.

163 Science 235:195–196; 1987. 14. Pankhurst, N. W.; Montgomery, J. C. Visual function in four Antarctic nototheniid fishes. J. Exp. Biol. 142:311 – 324; 1989. 15. Pankhurst, N. W.; Montgomery, J. C. Ontogeny of vision in the Antarctic fish Pagothenia borchgrevinki (Nototheniidae). Polar Biol. 10:419–422; 1990. 16. Pocket, S.; Macdonald, J. A. Temperature-dependence of neurotransmitter release in the antarctic fish Pagothenia borchgrevinki. Experientia 42:414–415; 1986. 17. Sullivan, C. W.; Palmisano, A. C.; Kottmeier, S.; McGrath, S. G.; Moe, R.; Taylor G. T. The influence of light on development and growth of sea-ice microbial communities in McMurdo Sound. Antarctic J. US 18:177–179; 1983.

/ eh08 2475 Mp 163 Wednesday Jan 08 04:13 PM EL–PB (v. 61, No. 2) 2475