Pacific Science (1975), Vol. 29, No.2, p. 211-21~ Printed in Great Britain

Function of the Dimorphic Eyes in the Midwater Squid

H istioteuthis dofl,eini I

RICHARD EDWARD YOUNG2

ABSTRACT: The squid Histioteuthis dofleini, like other members of the family Histioteuthidae, has a large left eye and a small right eye. The large eye points in a dorsal posterior direction while the squid typically orients at an oblique angle with the arms downward. The large eye, as a lesult, points vertically upward. The small eye appears to be directed ventrolaterally. This squid occurs primarily at depths of 500 to 700 m during the day where it is exposed to low levels ofdownwelling light. Presumably the large eye utilizes this faint downwelling light while the smaller eye utilizes bioluminescent light.

SQUIDS of the family Histioteuthidae exhibit a produce a less "grainy" image. Walls (1942: peculiar modification of the visual system. 210) stated, however, that in nocturnal verte­ During the larval stage, the eyes are normal in brates an enlarged eye is designed for greater size and shape. At the termination of the larval sensitivity rather than for resolution. Indeed, period, the left eye becomes atypical in shape one of the primary means of attaining high and rapidly enlarges relative to the right eye, sensitivity (retinal summation) is achieved at the diameter becoming neatly twice that of the sacrifice of acuity. The large eye may the right eye in juveniles and adults. Two provide a compromise between acuity and theories attempt to explain this peculiar devel­ sensitivity as apparently happens in geckos, opment. Voss (1967; Lane 1960: 110) suggested Sphenodon, and possibly owls (Walls 1942: 206), that the large eye functions when the animal but little can be said on this subject at present. is in the dimly lit waters of the deep sea, . Eye size in fishes and squids has only a whereas the normal eye functions in near­ marginal effect on the intensity of the retinal surface waters. Denton and Warren (1968) image due to the fixed relationship between suggested the exact opposite; that the large the size of the lens and the focal length of the eye is adapted for vision in near-surface waters, eye. The spherical lens is the only refractive and the small eye for vision in deep waters. structure in the eye (the cornea, when present, This idea, which is based on the presence of plays no role in focusing), and the lens shape pigment~ in the lens of the large eye which cannot be altered. The refractive index of the absorb ultraviolet radiation, will be examined lens, which is graded from the COle to the at the conclusion of this paper. Voss's sugges­ periphery, is constantly readjusted with growth tion that the large eye is an adaptation to the such that the size of the lens remains the only deep-sea habitat seems quite possible. factor affecting focal length (Pumphrey 1961). The value of a large eye to a deep-sea squid This fixed relationship of lens size to focal could be to increase visual sensitivity and/or length (retinal distance) is known as Matthies­ visual acuity. Compared to a small eye, a sen's ratio (the distance from the center of the large eye with a large retinal area may possess lens to the retina is 2.55 times the radius of a greater number of visual cells and thereby the lens). Walls (1942: 211) pointed out that, in such an eye, doubling the eye diameter

I This study was supported by National Science would double the diameter of the retinal Foundation grant no. GA-33659. Hawaii Institute of image. Thus, while more light is admitted in a Geophysics contribution no. 644. Manuscript received large eye, it is spread over a larger retina so . 21 January 1974. that illumination of the retina per unit area 2 University ofHawaii, Department ofOceanography, Honolulu, Hawaii 96822. remains the same. However, Denton and 211 212 PACIFIC SCIENCE, Volume 29, April 1975

Wanen (1957) and Clarke and Denton (1962) MATERIALS AND METHODS stated that, for seeing small spots of light, a All specimens were captured off the island large eye has an advantage. This advantage of Oahu in the Hawaiian archipelago at ap­ holds for point sources of light that will be proximately 158°18' W, 21°23' N over bot­ focused on single retinal cells and for small tom depths of 1,500 to 4,500 m. Two types of spots of light where the "grain" size of the trawls were used: a modified 3-meter Tucker retina becomes important. Thus, a large eye trawl and a 3-meter Isaacs-Kidd midwater with a retinal image covering an area four times trawl (IKMT). The Tucker trawl opens and that of a smaller eye will have an increased closes at the fishing depth; hence, capture of retinal intensity if the image falls on less than specimens during setting and retrieval of the four retinal cells. Therefore, while the retinal trawl (contamination) cannot occur. The intensity of small spots or· points of light opening-closing mechanism utilizes a mechani­ viewed by the eye is affected by eye size, the cal release that is activated by weighted messen­ retinal intensity oflarger objects viewed by the gers sent down the towing cable. eye is independent of the eye size. The IKMT is always open, and occasionally A large eye with a greater number of retinal specimens are captured while the trawl is cells, however, has a distinct advantage over being raised and lowered. This contamination a small eye in increasing sensitivity through is minimized by dropping the trawl as rapidly retinal summation. An all-rod vertebrate eye as possible and retrieving it with the ship can increase sensitivity about one millionfold moving slowly ahead. The net is pulled hori­ during dark adaptation, and one of the two zontally at 3 to 4 knots. Depth records for most important mechanisms involved is retinal both trawls were obtained with a Benthos summation (Tansley 1965). The high degree of time-depth recorder. retinal summation in nocturnal vertebrates (Tansley 1965: 51) further supports the im­ portance of this mechanism. Unfortunately, it RESULTS is not known whether or not retinal summation Description of Histioteuthis dofleini occurs in cephalopods. If it does occur, it ,will take place in the optic lobes (perhaps in Histioteuthis dojleini is similar in appeara~ce ·the "deep retina") where the axons from the to other members of the Histioteuthidae ex­ rc·tinal cells terminate. The complex structure cept that the arms are relatively long and the 9f these lobes prevents the detection of sum­ mantle small. The photophores of H. dojleini mation with simple anatomical techniques. exhibit a somewhat unusual distribution and Whatever the mechanisms involved, it does orientation for a midwater animal. The precise appear that a large eye in a midwater squid arrangement of these light organs is important would probably be very advantageous in, at to the subsequent discussion. On the mantle, least, increasing visual sensitivity. large photophores are concentrated on the If the large eye is an adaptation to the poorly anterior-ventral surface, with smaller and lighted waters of the deep sea and the small fewer photophores on the posterior and dorsal eye to the well-lighted surface waters, then surfaces. On the ventral surface of the head, knowledge of the precise habitats of these large photophores are closely spaced and squids could provide strong supporting·evi­ evenly distributed except for the left side, dence. The little information presently avail­ ventral and posterior to the large eye, where able on their habitats (Voss 1969; Roper and they are lacking. The lateral portions of the Young, in press), however, tends to contradict head anterior to the eyes also bear large photo­ such a relationship.. phores. The dorsal surface of the head bears, During a study of the vertical distribution of for the most part, only a few small photophores. pelagic cephalopods off Oahu, Hawaii, I have Seven large photophores are present near, but reexamined this problem based on information not at, the anterior-ventral edge of the large obtained on the vertical distribution and general left eyelid, whereas the smaller right eyelid biology of one species, Histioteuthis dojleini. .. possesses 17 large photophores tightly packed Histioteuthis dofleini-YOUNG 213

Lt. Eye Ace. Ret Main Ret. Opt. l. Supraes. Mass ...... ~ Tub. Eye

FIGURE 1. Histioteuthis dofleini. A, oblique section of the head ofH. dofleini passing through the visual axes of both eyes; B, outline of the drawing in A but with the outline of a tubular eye superimposed on the large left eye; C, dorsal-posterior view of H. dofleini in an aquarium; this view is presumably what one would see if one were looking vertically downward at a specimen floating in the water; D, lateral view showing the small right eye; E, lateral view showing the large left eye. The object in the photographs with the squid is a pair of 12-inch forceps. ABBREVIATIONS: Lt. eye, left eye; Ace. Ret., accessory retina; Main Ret., main retina; Opt. L., optic lobe; Supraes. Mass, supraesophageal mass; Rt. Eye, right eye; and Tub. Eye, tubular eye. 214 PACIFIC SCIENCE, Volume 29, April 1975

along the entire circular edge of the eyelid. posterior-dorsally to the main retina than The arrangement of reflectors and pigment on anterior-ventrally (Figure lA). Because the these latter photophores indicates that light eye is easily distorted by contraction of the from them does not enter the eye but passes head muscles during capture and fixation, it is anteriorly and somewhat laterally. Large not certain whether the accessory retina lies photophores are found on the aboral surfaces at the same distance from the lens as the main of all four pairs of arms, although they are retina or whether it is closer to the lens, as most numerous on the fourth (ventral) arms illustrated in Figure lA, B. and least numerous on the first (dorsal) arms. The orientation of the large left eye is All of the photophores face anteriorly. The atypical. Instead ofthe usual lateral orientation, skin ofH. dofleini contains many reddish brown the eye faces in a posterior-dorsal direction. chromatophores that can greatly alter the It is probably capable of limited movement. color of the animal. When the chromatophores In a living specimen of H. heteropsis off Cali­ contract, the animal looks silvery due to under­ fornia I have observed the large eye move lying iridophores; and when the chromato­ from a posterior-dorsal direction to a dorsal phores expand the animal becomes deep brown­ direction. Undistorted dead specimens of H. ish red. Chromatophores can also expand over dofleini invariably have this eye directed pos­ the silvery tissue of the photophores. The iri­ terior-dorsally, which is undoubtedly its more descent layer is most prominent on the ventral typical position. Because of the eye's orienta­ and lateral surfaces of the head, the anterior tion and large size, the head bulges laterally surface of the mantle, and the aboral surfaces and the margin of the eyelid is elliptical and of the arms. A weaker layer is present on the very large, passing around the lens and the dorsal surfaces of the head. Very little, if any, lateral wall of the bulbus of the eye. This iridescent tissue exists on the dorsal and lateral portion of the eye that is not covered posterior surfaces of the mantle. by the eyelid contains a layer of iridophores. The right and left eyes differ greatly from one another in size and shape (Figure lA). ,The right eye has a typical hemispherical shape Vertical Distribution .and a spherical lens. It differs from a typical The vertical distribution of H. dofleini is squid eye primarily in the structure of the presented in Figure 2. The symbols in the retina. The dorsal-posterior portion of the figure require some explanation. Tucker trawl retina is noticeably thicker than the anterior­ captures are represented by a vertical bar that ventral portion. The change in thickness is indicates the total range fished by the net while gradual, with the thickest portion being in the it was open. Within this range the net usually dorsal-posterior third and the thinnest in the fishes predominately within a narrow zone, anterior-ventral third. the midpoint of which is indicated by a dot. The spherical lens of the left eye has twice The IKMT also fishes primarily within a the diameter of the right one. In absolute narrow vertical range. The total range, of terms, the size of this eye is equally imp~essive. course, extends to the surface and, therefore, In a specimen of only 75 mm mantle length, is not represented in Figure 2. The probable the lens diameter is 15 mm. The eye does not depth of capture is determined from the hori­ have a hemispherical shape but has more the zontal phase of the tow in the same manner as shape of a truncated cone with a curved base. for captures from the Tucker trawl. Every The retina is divided into two portions of IKMT tow below 700 m during the day and different thicknesses. The main retina (thick below about 300 m at night passes through portion) is circular, nearly all of it being con­ the habitat of most of the population while the fined to the curved base ofthe cone and leaving net is being set and retrieved. In such circum­ most of the converging region of the eye free stances, some contamination is expected. I pf any retina. The accessory retina (thin have assumed that five specimens (represented 1?ortion) is continuous with the main retina on by the small dots in the figure) were captured l\ll sides, but covers a much broader area in this fashion. i Histioteuthis dofleini-YOlJNG 215

0 • 200 :.''''' .. • • • ,• • • 00 E 400 I 0 + I- 0+=, 0 Q.. , • w 600 ~ 0 a !O 000 0 ~ 0 , I , 0 , 6 b If : 800 0

1000

10 20 30 40 50 60 MANTLE LENGTH, mm

FIGURE 2. Vertical distribution of Histioteuthis dofteini off Hawaii. Each symbol represents a single capture. Large solid circle = depth of night captures; large open circles = depth of day captures; broken bars = depth range of opening-closing day tows; solid bars = depth range of opening-closing night tows; small solid circles = presumed contaminants.

The figure indicates that this species exhibits and squid (Young 1975) with tubular eyes. a diel vertical migration, moving upward Although larvae of H. dofleini live well above several hundred meters at night. During both the twilight zone in near-surface waters and the day and night, the larger animals occupy larger adults are found in the lower reaches of progressively greater depths. the twilight zone or occasionally below it, most juveniles and young adults occur within it and exhibit characteristics typical of many DISCUSSION squid living there, i.e., a layer of silvery The vertical distribution of H. dofleini clearly iridophores overlain by functional chromato­ indicates that its dimorphic eyes are not phores, and complex ventral photophores. adaptations to habitats with greatly differing Within this zone the intensity of down­ light intensities. Although this animal occurs welling light is over one hundred times in different day and night habitats, both greater than that passing upward (see Tyler habitats are characterized by very low light and Preisendorfer 1962: 423). Determining levels. the typical orientation of H. dofleini within The daytime habitat of most H. dofleini is a this strongly directional radiance pattern is zone of low light intensity, yet a zone where critical to understanding the functions of the light plays a critical role in the ecology of eyes. Clarke, Denton, and Gilpin-Brown (1969) many of the inhabitants. This twilight zone have shown that a closely related squid, H. from about 400- to 700-m depth off Hawaii reversa, is neutrally buoyant. H. dofleini also corresponds to the habitat of most half-red appears to be nearly neutrally buoyant in an shrimp (Foxton 1970; ]. Walters, personal aquarium. Although the animal tends to communication), to the habitat of most rotate to a position with the mantle downward animals bearing complex ventral photophores when motionless, it is probably capable of (Foxton 1970, Young 1973) and to the habitat orienting in almost any direction with only a of most fish (T. Clarke, personal communica­ slight assist from the fins and funnel. The tion; S. Amesbury, personal communication) normal orientation of this species can be 216 PACIFIC SCIENCE, Volume 29, April 1975 deduced from the distribution of photophores the ability to see details above the visual on the head, mantle, and arms. These photo­ threshold by increasing the signal-to-noise phores point in an anterior-ventral direction ratio; actual retinal stimulation ("signal") could relative to the longitudinal axis of the body. be distinguished from fluctuations in cell If the photophores are used in ventral counter­ activity (noise) by analyzing retinal stimulation shading (i.e., elimination of the silhouette coincident on the two retinas. when viewed against the downwelling surface Although most authors (e.g., Walls 1942, light), as are similar photophores in many Tansley 1965) have suggested that tubular other midwater animals (Clarke 1963; Foxton eyes represent large eyes, i.e., they correspond 1970; Denton, Gilpin-Brown, and Wtight1972; to the central cores of large eyes, this has not Young 1973), they then must be directed down­ been rigorously demonstrated. The tubular ward. In order to direct the photophores down­ eyes in fishes and cephalopods could represent ward the squid must be positioned with the the reduction of normal-sized eyes into com­ body axis at an angle of about 45° from the pact spaces for parallel alignment and binocular horizontal with the mantle uppermost. In this vision. However, the semitubular eye of H. orientation the large eye looks upward in the dofleini is clearly not just a normal-sized eye direction of maximum light intensity. in a somewhat compact form. Rather it is The brge eye of H. dofleini approaches a nearly twice the size of its right counterpart. tubular eye in shape and has certain functional The arrangement in H. dofleini demonstrates relationships to tubular eyes. In nearly all the problem of having large upward-directed tubular-eyed species, the eyes seem to be eyes. Even though the eye does not have the directed either dorsally on a horizontally full normal shape in this species, it still grossly positioned animal (e.g., Opisthoproctus) or distorts a very large head. By analogy, this anteriorly on a presumably vertically oriented arrangement in Histioteuthis suggests that tubu­ animal (most species with anteriorly directed lar eyes of fish are indeed large eyes in a com­ eyes may orient vertically, as has been indi­ pact form designed for viewing the vertically cated for Gigantura and Srylephorus [Bruun 1957]. downwelling light, and that binocularity, However, the fish Winteria may be an excep­ therefore, is not the only factor involved. tion. Unfortunately the evidence concerning The small eye of H. d?fleini has an anterior orientation in midwater animals is meagre.) tilt. An additional ventral tilt results from the The large histioteuthid eye is also directed tilt of the head imposed by the large right eye. upward and its visual field completely includes This latter tilt also explains the asymmetrical the vertical visual field of a tubular eye; yet, arrangement of the photophores on the ventral unlike the tubular eye, it maintains a broad surface of the head. When the animal is in its lateral field of view (Figure 1E). presumed typical orientation, the small eye The probable function of a tubular eye has tilts slightly downward. Therefore, while the been examined by Munk (1966) and others. large and presumably more sensitive eye points Munk demonstrated that the tubular eye is upward in the direction of maximum light equivalent to the central core ofa hemispherical intensity, the smaller eye, directed laterally eye. Fishes and cephalopods with tubular eyes and ventrally, points in a direction of very have the optical axes of their eyes parallel or low light intensity. The latter eye probably nearly parallel. The compact configuration of receives less than 5 percent of the downwelling the tubular eyes facilitates the parallel orienta­ illumination that the upward-looking eye re­ tion and thus binocular vision. Brauer (1908) ceives (see Tyler and Preisendorfer 1962: 423). suggested that binocular vision results in These circumstances suggest that, whereas the better judgment of distances, whereas Weale large upward-looking eye detects downwelling (1955) suggested that it results in lowering of surface light, the small eye does not; rather, the visual threshold. (Pirenne [1967] stated it detects bioluminescent light. Further, the that binocular vision lowers the visual thres­ compact arrangement of photophores around hold in humans by 20 percent.) Fremlin (1972) the smaller eye, combined with the modifica­ suggested that binocular vision may increase tions of the retina, suggests that counter- Histioteuthis dofleini-YOUNG 217 shading is not the only function ofthese organs. has been previously suggested by several These ocular photophores are ideally located authors. to produce a strong beam of light that would 5. The small right eye may function primarily illuminate the portion of the environment that in the detection of bioluminescent light is surveyed by the thicker portions of the and the photophores that surround the lens retina. In other words, these photophores may may function as a searchlight. function as a searchlight. 6. The large left eye functions primarily during the day in detecting downwelling light. Lens Pigments Denton and Warren (1968) suggested that ACKNOWLEDGMENTS the large eye of Histioteuthis functions in near­ I wish to thank Andrew Packard, University surface waters because of the light-absorbing Medical School, Edinburgh, and John Walters, characteristics of the lens. They found that in University ofHawaii, for reading and comment­ H. meleagroteuthis the lens of the small eye is ing on the manuscript. I also thank John transparent to light down to about 310 nm, Walters for taking the photographs in Figure whereas the lens of the large eye always ab­ 2, and Thomas Clarke, University of Hawaii, sorbs the near ultraviolet and sometimes blue for supplying most of the specimens taken in light. These authors pointed out that such open nets. features are characteristic of surface-dwelling fishes and squids. However, catch records LITERATURE CITED demonstrate that Histioteuthis spp. do not nor­ BRAUER, A. 1908. Die Tiefsee-Fische. 2. Anat. mally occur in near-surface waters during the Teil. Wiss. Ergebn. 'Valdivia' 15: 1-266. day-time and the previous discussion has BRUUN, A. F. 1957. Deep sea and abyssal attempted to demonstrate that the large eye is depths. Mem. geol. Soc. Amer. 67: 641-672. adapted for vision under conditions of low CLARKE, G. L. 1971. Light conditions in the light intensity. In near-surface species, an sea in relation to the diurnal vertical migra­ ultraviolet-absorbing lens may protect the tions of animals. Pages 41-50 in G. B. retina from damage (Denton and Warren Farquhar, ed. Proceedings ofan international 1968) or may improve visual acuity by re­ symposium on biological sound scattering ducing chromatic aberration (Wald and Griffin in the ocean. Maury Center for Ocean 1947). The reason for the ultraviolet-absorbing Science, Washington. pigments in the large lens of Histiotheuthis CLARKE, G. L., and E. J. DENTON. 1962. Light remains a mystery. and animal life. Pages 456--468 in M. N. Hill, ed. The sea. Vol. 1. John Wiley & Sons, Interscience, New York. SUMMARY CLARKE, M. R., E. ]. DENTON, and ]. B. 1. Histioteuthis dofleini lives primarily in the GILPIN-BROWN. 1969. On the buoyancy of twilight zone (approximately 400 to 700 m squid of the families Histioteuthidae, Octo­ depth) during the day and migrates upward poteuthidae and Chiroteuthidae. J. Physiol. several hundred meters at night. 203: 49-50. 2. This squid has a large left eye with a semi­ CLARKE, W. D. 1963. Function of biolumines­ tubular shape and a small right eye with a cence in mesopelagic organisms. Nature, typical hemispherical shape. Lond. 198: 1244-1246. 3. H. dofleini probably orients at an oblique DENTON, E. ]., ]. B. GILPIN-BROWN, and P. G. angle in the water so that the large eye is WRIGHT. 1972. The angular distribution of directed vertically upward while the small the light produced by some mesopelagic eye is directed ventral-laterally. fish in relation to their camouflage. Proc. 4. The disparity in the size of the two eyes roy. Soc., B, 182: 145-158. suggests that tubular eyes in other mid­ DENTON, E. ]., and F. ]. WARREN. 1957. The water animals are indeed enlarged eyes, as photosensitive pigments in the retinae of 218 PACIFIC SCIENCE, Volume 29, April 1975

deep-sea fish. J. Mar. bioI. Ass. U.K. 36: TANSLEY, K. 1965. Vision in vertebrates. 651-662. Chapman & Hall, London. ---. 1968. Eyes of the Histioteuthidae. TYLER, J. E., and R. W. PREISENDORFER. 1962. Nature, Lond. 219: 400-401. Transmission of energy within the sea. FOXTON, P. 1970. The vertical distribution of Pages 397-451 in M. N. Hill, ed. The sea. pelagic decapods (Crustacea: Natantia) Vol. 1. John Wiley & Sons, Interscience, collected on the SOND cruise 1965. J. Mar. New York. bioI. Ass. U.K. 50: 939-960. Voss, G. L. 1967. The biology and bathymetric FREMLIN, J. 1972. How stereoscopic vision distribution of deep-sea cephalopods. Stud. evolved. New Sci. 56: 26-28. trap. Oceanogr. 5: 511-535. KAMPA, E. M. 1971. Photoenvironment and Voss, N. A. 1969. A monograph of the Cepha­ sonic scattering. Pages 41-50 in G. B. lopoda of the North Atlantic: the family Farquhar, ed. Proceedings ofan international Histioteuthidae. Bull. mar. Sci. 19(4): 713­ symposium on biological sound scattering 867. in the ocean. Maury Center for Ocean WALD, G., and D. R. GRIFFIN. 1947. The Science, Washington. change in refractive power of the human LANE, F. W. 1960. Kingdom of the octopus. eye in dim and bright light. J. opt. Soc. Sheridan House, New YOlk. 300 pp. Amer. 37: 321-336. MUNK, O. 1966. Ocular anatomy ofsome deep­ WALLS, G. L. 1942. The:vertebrate eye. Bull. sea teleosts. Dana Rep. 70: 1-62. Cranbraok Inst. Sci. 19. 785 pp. PIRENNE, M. H. 1967. Vision and the eye. WEALE, R. A. 1955. Binocular vision and deep­ Chapman & Hall, London. sea fish. Nature, Lond. 175: 996. PUMPHREY, R. J. 1961. Concerning vision. YOUNG, R. E. 1973. Information feedback Pages 193-208 in J. A. Ramsey and V. V. from photophores and ventral counter­ Wigglesworth, eds. The cell and the organ­ shading in mid-water squid. Pacif. Sci. ism. At the University Press, Cambridge. 27(1): 1-7. ROPER, C. F. E., and R. E. YOUNG. In press. ---. 1975. Transitory eye shapes and the The vertical distribution of pelagic cephalo­ vertical distribution of two midwater squids. pods. Smithson. Contr. ZooI. Pacif. Sci. 29(3): 243-255.