The organization of spatial vision in the juvenile lemon shark (Negaprion brevirostris)

MISSING IMAGE

Material Information

Title:
The organization of spatial vision in the juvenile lemon shark (Negaprion brevirostris) retinotectal projection, retinal topography, and implications for the visual ecology of sharks
Physical Description:
vii, 130 leaves : ill. ; 28 cm.
Language:
English
Creator:
Hueter, Robert Edward, 1952-
Publisher:
s.n.
Publication Date:

Subjects

Subjects / Keywords:
Sharks -- Sense organs   ( lcsh )
Vision   ( lcsh )
Zoology thesis Ph. D
Dissertations, Academic -- Zoology -- UF
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1988.
Bibliography:
Bibliography: leaves 118-129.
Statement of Responsibility:
by Robert Edward Hueter.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001076289
oclc - 19045338
notis - AFG1048
System ID:
AA00002139:00001

Full Text












THE ORGANIZATION OF SPATIAL VISION
IN THE JUVENILE LEMON SHARK (NEGAPRION BREVIROSTRIS):
RETINOTECTAL PROJECTION, RETINAL TOPOGRAPHY,
AND IMPLICATIONS FOR THE VISUAL ECOLOGY OF SHARKS


ROBERT EDWARD


HUETER


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
















ACKNOWLEDGMENTS


many


problems


of conducting


physiological


research


on living


sharks


could


only


have


been


overcome with


generous


assistance


many


people.


first


extend


my deepest


appreciation


and gratitude


major


professor,


Dr. Horst


Schwassmann,


one of the original


"mappers"


from


a time


when


technology


was second


to technique.


His initial


skep-


ticism about


proj ect


was


always s


tempered


with


an undercurrent


strong


support


which


soon


be came


enthusiastic


encouragement


as the


pro-


ject


progressed.


have


learned much


from


him,


not only


in electrophys-


iology,


histology,


sensory


biology,


but also


in appreciating


historical


perspective


of scientific


endeavors--not


to mention


many


splendors


also


the fish


express


eye.


sincere


For all of these


appreciation


things


to Dr. Paul


am deeply

Linser a


grateful.


nd Dr.


William Dawson


my supervisory


committee.


Both


them generously


opened


their


laboratories


to me


so that


could


success


fully


complete


this


project.


Dr. Linser' s


support


of my work


during


tenure


at the


Whitney


Laboratory was


extraordinary,


literally


keeping


this


project


afloat,


and to him


am deeply


grate


ful for that


support


as well


as his


expertise,


his patience,


and his friendship.


Much


same


also


applies


to Dr. Dawson,


whose


vision


laboratory


opened


y own eyes


_ w_









Dr. David


Evans


and Dr. Carter


Gilbert,


two other members


my committee,


have


been


among


most


influential


teachers


uni-


versity


education,


and I


thank


them


for their


knowledge,


their


inspir-


action


example,


and their


friend


ship.


To the following,


expert


am deeply


in the collection and


indebted


maintenance


for their


of sharks


assistance


in captivity.


Frank Murru and


Chittick


of Sea World


of Florida


supervised


the col-


lecting


my experimental


animals,


furnished


transport


equipment


provided


advice


on captive maintenance


of the sharks.


Billy


Raulerson


the Whitney

ogy departme


Laboratory,


and Robert


in Gainesville


helped


ngman

solve


and Charlie J

the problems


abaly


at the zool-


of keeping


sharks


in limited


facilities.


Special


thanks


are extended


to the following


people


at the Whitney


Laboratory


support


for their


and technical


assistance:


help


Drs.


Margaret


Mike


Perkins


Greenberg


and Barry


for technical


Ache


and moral


sup-


port;


and Lynn Milstead


and Jim Netherton


for illustration


and photo-


graphic work.


Back


in Gainesville,


thank


Farmer


for technical


distance,


and Minnie


Hawthorne


deserves


special mention


for her help


the pursuit


of the perfect


retinal


wholemount.


Financial


assistance


for this


project


was provided


the following


sources:


the Lerner-Gray


Fund


for Marine


Research,


a Sigma Xi


Grant-in-


Aid,


and the Department


of Zoology


at the University


of Florida.


Finally,


personal


thanks


are extended


friend


and colleague


Dr. Joel


Cohen


for his


constant


encouragement,


expertise,


and special


S* ^ +'ln r t


rFI- n A '


as-


c- rll l 1 ,Ir I1


r1 ^V\ a 1r crHQT I


-1K <"' 4--< Yln


Llrl ^T-T- /"/-

















TABLE


ACKNOWLEDGMENTS

ABSTRACT. .

INTRODUCTION.


OF CONTENTS


S. . .


. . .


Physiological
Topography of
The Midbrain
Retinotectal


Vertebrate


Vision
Vision


Visual Pathway


Proje


action


S 4 . 1


I .


Patterns


Retinal
Topography
Tectal A
Retinal


Topography.
of Spatial


anatomy


S . . .


Vision


and Physiolo


in Sharks. .
gy S S . .


Topography.


Studies


The Juvenile


Biology
Visual


Spatia
Lemon


Visual


rformance


Shark.


the Lemon


stem


Shark


Features


Objectives


MATERIALS


the Study


METHODS


Animal


Collection


and Maintenance


Colle


action.


Transport


and Maintenance


teams


Food


and Feeding.


Preparation


for Experiments


Retinotectal Mappin


Experimen


ts. S S S S S S


Surgical


Preparation.


Mappin
Photic


Chamber
Stimuli.


* S S . . .
* . . . 0 . .


Electrode
Mapping
Histology
Fixation


Retinal


es


and Instrumentation.


Procedure . . .
. . .


Wholemounts









RESULTS.


Retinotectal Mapping


Responsive


Layer


and Receptive


S. . 57


Retinot


ectal


General


Topogr


Comp


Topography.


proj section


aphic


osite


pattern


maps


maps


Retinotectal magnification


factor


S 84


Retinal


Topography


. 85


Photoreceptor


and Ganglion


Cell


Identifi


cation


. . 85


Cone


and Ganglion


Cell Dens


. 95


DISCUSSION


The Retinotectal


Pathway


in the Lemon


Shark.


Nature
General


the Tectal


Pattern


Units


and Functional


Significance


Topographical
Nonlinearity


Anatomical
Implications


Organization


of Vision


in Retinotectal


Basis


in the Lemon


Shark.


Topography.


in the Retina


for the Visual


Ecolo


of Sharks.


Function


of the Visual


Streak.


Notes


on Visual


Behavior


in the Juvenile


Lemon


Shark


Visual


Niches


of Sharks


SUMMARY


AND CONCLUSIONS.


REFERENCES


BIOGRAPHICAL


SKETCH.


Fields
















Abstract


of Dissertation


of the University


of Florida


Presented


to the Graduate


in Partial


Fulfillment


School
of the


Requirements


for the Degree


Doctor


of Philosophy


THE ORGANIZATION


OF SPATIAL


VISION


IN THE


JUVENILE LEMON


SHARK


(NEGAPRION


BREVIROSTRIS)


RETINOTECTAL


PROJECTION


RETINAL


TOPOGRAPHY,


IMPLICATIONS


THE VISUAL


ECOLOGY


OF SHARKS


Robert


Edward


Hueter


April


1988


Chairman:


Major


Horst


Department:


Schwassmann


Zoology


an effort


to understand


the organization


of spatial


vision


its role


in the visual


ecology


of sharks,


the retinotectal


projection,


retinal


cone


distribution,


and retinal


ganglion


cell


distribution were


mapped


in the juvenile


lemon


shark.


Retinotectal mapping was


electro-


physiolo


simulated


ically

the 1


conducted

eft visual


in live

field.


sharks

Tectal


using

unit


a perimetry


responses


device


evoked


that

pho-


tic stimuli


were


recorded


in the contralateral


optic


tectum


an aver-


age depth


below


the tectal


surface.


overall


projection


pattern


reveals


an orderly


point-to-point


relationship


such


that


dorsal


visual

eral t


field


ectum,


projects

rostral


to medial


field


tectum,


proj ects


ventral


to rostral


field

tectum


projects to lat-

and caudal field


I _


__


I_









Retinal


cell


topography was


mapped


using


a retina wholemount


tech-


unique.


The retinotectal


projection


pattern,


cone


distribution,


gan-


glion


cell


distribution


are all organized


into


a prominent


visual


streak,


a horizontal


band


of proportionately


greater


retinal


cell


density


retinotectal magnification.


This


streak


is located


within


about


above


and 15


below


the horizontal


meridian


in the visual


field.


Three


times


more


tectum is


devoted


to vision


in the streak


than


to peripheral


vision,


and receptive


fields


of tectal


units


are smaller


and more


regu-


lar in shape


inside


the streak


than


in the periphery.


Ganglion


cell


cone


densities


increase


from


less


than


anglion


cells/mm


less


than


500 cones/mm


peripherally


over


1500


ganglion


cells/mm2


about


6500


cones/mm2


inside


the streak.


The visual


streak


of the juvenile


lemon


shark


conforms with


terrain


theory


of visual


streak


function.


According


to this


theory,


streak


enhances


spatial


vision along


the visual


horizon in


animals whose


habitats


are dominated


a two-dimensional


horizontal


terrain.


In the


juvenile


lemon


shark,


the locomotory mode


constant


patrolling


over


the benthos may


further


add to the adaptive


value


of the visual


streak


in this


species.















INTRODUCTION


Physiological


Ecology


of Vision


sensory


systems


of animals


constitute


the selective


channels


through


which


animals


perceive


their world


and gain


information


about


their


pass


environment.


through


All that


the filters


an animal


learns


one or more


of its surroundings


senses.


To make


must


best


use of available


resources


, be they


food,


protection


from


predators,


reproductive


opportunities


an animal's


sensory


systems


must


no less


exquisitely molded


the forces


of natural


selection


than


its locomo-


tory


design,


respiratory


rate,


or reproductive


strategy.


From


this


premise,


it follows


that


a study


of the


ways


in which


sensory


systems


are attuned


to the environments


in which


they


operate


yields


at least


two interrelated


kinds


of insights


into


animals


' lives.


The first

particular


of these

species.


is the bod

Careful


significant


examination


sensory


an animal


stimuli


s sensory machin-


can reveal


the key


stimuli


selected


for by


that


species.


The dis-


coveries


of bombykol


olfactory


receptors


in the


antennae


of the silk-


worm moth


Schneider,


1971)


movement-sensitive


ganglion


cells


in the


frog


retina


(Lettvin


et al.,


1959),


and the ultrasensitivity


of ampul-








and listened


for in


an animal'


daily


life.


They present


a picture


the animal


itself


perceives


its own


environment.


The second


insight


gained


these


studies


a comparative


under-


standing


of how


species


' senses


are designed


to accommodate


the total


range


environmental


information


available


to them,


perhaps


leading


more


refined


explanations


of species


success.


Both


types


of insights


are concerned with


the linka


between


animal


and environment,


and with


the utilization


sensory


information


as an environmental


resource.


With


the diversification


sensory


systems


through an


evolutionary pro-


cess


of adaptive


radiation


(Ali,


1978),


sensory


resources


are subdivided


among


species,


just


as food


and living


space


are partitioned within


eco-


logi


cal niches.


In attempting


to comprehend


the remarkable


diversity


of adaptations


and complexities


found


in the


sensory


systems


animals


it is useful


to consider


concept


of the "sensory


niche.


" A


sensory


niche


can be


defined


as the total


range


of environmental


information


that


is detected


and monitor

ecological


the combined


expression


sensory


of the sampling


systems


strategies


a species.

used by a


It is


species


selecting


out specific


sensory


information


relevant


to species


survival.


The boundaries


of the niche within


a particular


sensory


environment


set by


the filtering


characteristics


of the specific


sensory


systems


involved,


matic


such


vision.


as frequency

In this way,


response


sensory


in hearing or


niche


chromatic


represents


achro-


an animal's


total


perceived


environment,


which


a subset


of its broader physical


environment.


are









1987).


Yet it


seems


apparent


that


research


on the functional


perform-


ance


a sensory


system


can truly


a study


of the "sensory


physiolo-


gical


ecology"


a species.


a viable,


recognized


subdiscipline


within


biology


such


a field


unfortunately


does


not yet


exist


, although


concept


of "sensory


ecology"


has been


defined


was


the subject


at least

limited


one major

standpoint


symposium


strictly


(Ali,


1978).


neural


But from


elements


the somewhat


in sensory


systems


more

- we do


have


a viable


concept


of second


choice


in pursuing


such


research:


neuroethology,


the neural


basis


of animal


behavior.


Although


neuroetho-


logy


is historically


the field


has today


rooted

grown t


in studies


o include


of brain


research


physiology


with


(Ewer


implications


t, 1980),

not only


for animal


behavior


in a pure


ethological


sense,


but also


for animal


life


history


and ecology


as well.


The role


of vision


in the life


of animals


has received


special


attention


from


neuroethologists,


because


at least


two factors.


of these

a natural


is that


vision


curiosity


about


is the dominant


how other


sense


species


in humans,


see the world


so we have


and if their


vision


is "as


good"


or "better"


than


our own.


The second


factor


involves


the special


characteristics


of the vis-


sensory


channel.


Unlike


some


of the other


senses,


vision


operates


with


discrete


stimuli


having


extremely


precise


temporal


and spatial


properties,


and these


features


are specifiable


and quantifiable


high


degree.


This


increases


the information-carrying


capacity


of the


visual


channel,


and,


in some


respects,


it helps


to simplify


the work


ncnT rnorhnl nr4r c t c, f i 0 i m n


nn anlma1


7u77 Qi nn -


-in in *n+*Qt-TT- ^^ -I


Q-vn T ri mon + Q









Topography


of Vertebrate


Vision


The Midbrain


Visual


Pathway


emergent


theme


property


a topographic


in the design


representation


many


of the


sensory


sensory


systems


umwelt


is the


at points


within


the central


nervous


system of


an animal


(Ulinski,


1984)


For the


somatosensory


takes


system


on the form of


bolically


represented


or sense


of touch,


a somatotopic


the bizarre


for example,


proj section,


which


homunculuss"


this "map"

in humans


relating


skin


concept


sym-


surfaces


to projection


sites


on the cerebral


cortex


(Penfield and


Rasmussen,


1950).


For the visual


system


this


type


of topographic map


is called


retinotopic


projection,


relating


points within


the visual


field


to their


corresponding


representations


in the retina


and brain.


As with


matotopic m

retinotopic


lap,


the proportional


projection


pattern


representation

conforms with


sensory


the relative


areas i

spatial


n the

impor-


tance


of zones within


the visual


field


to the animal.


In other words,


the map


indicates what


areas


of an animal


s visual


field


are being


looked


at with


a greater


investment


of neural


components,


with


the impli-


cation


that


the details


of such


areas


are


somehow


of greater visual


portance


to the animal.


a sense,


the retinotopic map


shows


the animal


itself


sees


the visual


space


around


into


this


central


proj section


the visual


world,


then,


to gain


some


comprehension


of the handling


of spatial


information


in the


-r


*


-------AA- -- ---- -- -


so-


A


7t ^ _-


I


I


1


_,_


- A- --


^ _


Am









Beginning


approximately


two decades


, a major


theme


in visual


science


took


shape


in which


the central


pathways


of vision


verte-


brates


are divided


into


primarily


two different


visual


systems


(Schnei-


der,


1969).


The first


pathway--and


probably


the "older"'


two,


evolutionarily


speaking--projects


from


the retina


to the mesencephalic


optic


tectum


or midbrainn


roof"


(superior


colliculus


in mammals)


with


postsynaptic


continuations


to other


sites


in the CNS,


including


spinal


cord


and medulla


retinotectal


(myelencephalon)


projection


is the dom


and thalamus

inant visual


(diencephalon).


pathway


This


in nonmammalian


vertebrates.


The second


pathway,


most


highly


developed


in mammals


, projects


from


the retina


to the diencephalon


(lateral


geniculate


nucleus),


where


synapses


with


pathway


continuations


to the telencephalon


striate


cor-


tex in mammals).


Many


efferent


connections


are made


from


visual


centers


in the


cortex,


including


pathways


back


to the diencephalon


mesen-


cephalon.


In both


pathways


, the primary


crossing


of optic


nerve


fibers


to the


opposite


side


(contralateral


projections)


occurs


between


retina


and mes-


encephalon


or diencephalon.


the first


pathway,


this


decussation


fibers


between


retina


and mesencephalon


is complete


in all nonmammalian


vertebrates.


In other words,


primary


retinotopic


projection


lower vertebrates


mammals,


is found


the decussation


in the contralateral


fibers


optic


in the dominant


tectum.


pathway


But in


between


re-


tina


and lateral


geniculate


nucleus


partial,


resumably


a modifica-


A- .


-I rnnr \


/T- -- n-_


. *


S11


*r I -_* -









1946,


the groundwork was


being


laid


to probe


this


dichotomy.


Working


with


cats,


Hess


et al.


(1946)


used


electrical


stimulation


super-


ior colliculus


delivering


to elicit

physiologic


ehavioral


al levels


correlates


of midbrain


stimulation


a sing


function.

le superior


colliculus


in the


cat brain,


Hess


et al. observed


rapid,


distinctive


head


eye turning movements


symmetrically


opposite


to the side


stim-


ulated.


This


spatial


response


was interpreted


as being


a function


decussating


visual


pathways


projecting


from


eyes


to the contrala-


teral


superior


colliculi.


perhaps


the most


important


aspect


of the work


of Hess


et al.


(1946)


was


that


midbrain


they


attempted


pathway


terms


to explain


a reflex


the functional


response,


significance


the "visuelle


of this


Greifre-


flex,"


or visual


grasp


reflex.


According


to Hess


the midbrain


pathway


provides


for a rapid


reflex


to the perception


objects


in motion,


that


eyes


and head


are abruptly


turned


to "grasp,


or fixate


on, objects


that


appear


instantaneously


in the peripheral


visual field.


This


reflex


is mediated


corresponding


tectospinal


pathways


projecting


to motor


nuclei

cation


connecting with

of the findings


ocular,

of Hess


neck,

et al.


and

was


upper

that,


chest muscles.


at least


The impli-


in mammals,


midbrain


pathway


controls


primary


orienting movements


towards


just-


detected


objects


in the contralateral


visual


field,


especially moving


objects


at the periphery.


The work


of Hess


et al.


was


followed


stimulation


studies


on the


optic


tecta


of lower


vertebrates.


Both Akert


(1949)


and Leghissa


(1951)


c,*..A4 innCa0rrnl


4 ,* nw i t'1 a


I tr\ +-Q-f-a


~c'+'t/^- n<"


nnnnl*


A-F 4'-! ^1-tr


r*E e\ c? T^aE- c c


n









stimulation


of the contralateral


optic


tectum.


Akert


stated


more


expli-


citly


the function


of the reflex:


an object


appearing


in the periphery


of the visual


field


can be scrutinized


with


optimum


acuity


through


voluntary


change


of ocular


position,


to align


target


with


higher


acuity


zones


in the central


visual


field.


Akert


furthermore


described


systematic


correlation


between


coverage


areas


of visual


space


on the


retina


and the representation


of those


areas


on the fish


tectum,


with


experimental


histological


tra-


cing


of optic


fibers.


The correspondence


sensory


afferents


with


motor


efferents


in the


tectum,


Akert


said


, pointed


to a functional


rela-


tionship


between


the perception


an object


in visual


space


sponsive


turning movements


toward


such


an object.


Thus


Akert


elabor-


ated


on Hess


et al.'s


theme


not only


extending


it to the midbrain


visual


pathway


of vertebrates


in general,


but also


investigating


the functional


, topographic


projection


patterns


of that


pathway.


Leghissa


(1951)


explored


same


correlation


between


specific


movements


and contralateral


tectal


stimulation


in goldfish


(Carassius


auratus).


He also


found


a correspondence


between


stimulated


tectal


areas


and resulting


shifts


in the visual


field


directed


move-


ments.


Later,


Ewert


(1967)


demonstrated


that


toads


(Bufo bufo)


also


make


turning movements


in response


to contralateral


tectal


stimulation.


Thus,


similar


functions


were


uncovered


for midbrain


pathways


in at


least


fishes


, amphibians,


and mammals.


From


the studies


of Hess,


Akert,


Leghissa,


Ewert


and others,


thPrPfnr .


pmperced


a ppneral zed


themP


!


the vertehrate


midbrain


oath-


re-


,









functional


basis


to the previously


reported


retinotectal


projection


patterns,


explored


anatomically


beginning with


Lubsen


(1921)


and elec-


trophysiologically


beginning with Apter


(1945).


This


functional


basis


was rooted


in the discovery


of responsive


orienting


movements,


which


could


be elicited


stimulation,


either natural


or experimental,


tectospinal


pathways.


Further


that,


studies


in general,


precisely


ordered


on various


the midbrain


retinal


vertebrate

tectum or s


projection


groups


superior


preserving


have


demonstrated


colliculus


the topographic


receives


order


of the retina


and visual


field,


and the midbrain


structures


furthermore


send


efferent


connections


directly


to brain


stem


and spinal motor


cen-


ters


(Ingle,


1973).


Retinotectal


Projection


Patterns


Although


use of electrophysiological


mapping


techniques


trace


neural


pathways


dates


back


at least


to the 1870's,


they


remain


valuable


approach


to the understanding


of whole macroscopic


patterns


neural


projections,


especially when


combined with


neuroanatomical


tech-


niques


chlag,


1978).


Vision


researchers


seized


upon


this


methodolo


especially


in the 1950'


and 1960'


when a


host


reports


appeared


the wake


Apter


s mapping


of the retinotopic


projection


on the


super-


ior colliculus


cat (Apter,


1945).


Since


then,


precise


retino-


topic


proj section


patterns


as revealed


electrophysiological


mapping


have


been


described


ecies


of bony


fishes


(Jacobson


and Gaze,


1964;


Schwas smann


and Kruger,


1965a,b


Schwassmann,


1968;


Schwassmann









Whitteridge,


1954),


and mammals


(Hamdi


and Whitteridge,


1953;


Siminoff


et al. ,


1966;


Pettigrew,


1986a).


two classes


of vertebrates


missing


from


this


survey


are the


agnathans


(lampreys


and hagfishes)


and the chondrichthyans


(sharks,


skates


rays,


and chimeras).


Visual


evoked


potentials


have


been


recorded

et al.,


from

1966)


the mesencephalon


but precise


of lampreys


retinotectal


(Veselkin,


topography


1966; Karamian

not been mapped.


This


not so surprising


considering


the minute


size


of the lamprey


brain.


But what


less


understandable


is the complete


omission


chondrichthyan


species


from


the list,


in spite


of the large


size


accessibility


of the elasmobranch


tectum,


and the strategic


phylogenetic


position


of the


group


in vertebrate


evolution


(Bullock,


1984).


entire


class


has simply


been


ignored


as the field


has moved


on to other


topics,


such


as retinotectal


specificity


regeneration


g. Sharma,


1972;


Schmidt


, 1982)


and integration


of multimodal


input


in the


vertebrat

a number


mesencephalon.


of species


The latter


including


has been


elasmobranchs


the subject


(Platt


of study


et al., 1974;


Schweitzer,


1986),


teleosts


(Bastian,


1982)


reptiles


(Hartline


et al.


1978;


Stein


and Gaither,


1981)


and birds


(Knudsen,


1985).


The elasmobranchs


notwithstanding,


the basic


patterns


of retino-


tectal


projection


and retinal


structure


in fishes


have


been


well


studied,


as fishes


have


long


served


as models


for visual


science.


As the oldest


general


group


of vertebrates,


the fishes


were


first


to evolve


verte-


brate


eye,


and,


judging


the remarkable


constancy


structure


in the


__ 1









high


acuity


zones


such


as foveas


and visual


"areas,


and mechanisms


light


and dark


adaptation


(Walls,


1942).


As previously


stated,


the primary


topographic


of the visual


pathway


between


retina


and brain


in fishes


is the retinotectal


proj ec-


tion,


and fish


retinotectal


topography


has been


most


precisely


inves-


tigated


using


extracellular microelectrode


recordings


at the


tectum


during


photic


stimulation


of the retina


(S chwas smann,


1975


Vane


1983).


Where


equal


areas


of the visual


field


have


proportionately


equal


representation


on the tectal


surface,


the retinotectal


projection


pat-


tern


is called


linear


(Schwassmann,


1968).


A linear


projection


pattern


is indicative


of uniform,


unspecialized


spatial


vision,


in which


no par-


ticular


region within


an animal's


visual


field


receives


greater


atten-


tion


to detail


than


other


region.


Such


a linear


pattern


has been


found


terus


in four


salmoides;


species


of freshwater


bluegill,


Lepomis


teleosts:


macrochirus;


largemouth


carp


bass,


Cyprinus


Microp-


carpio;


and goldfish,


Carassius


auratus


(Schwassmann


and Kruger,


1965a;


Jacob-


son and Gaze,


1964).


On the other


hand,


nonlinear patterns


indicating


regions


in the


visual


field


receiving, magnification


of the retinotectal


projection--


that


portion


a proportionately


of the visual


greater


field--are


amount


found


tectum


devoted


in vertebrates with


a given


some


type


heterogeneity


in their


spatial


vision.


These


animals


are specialized


for potentially


greater visual


acuity


along


certain


lines


of sight,


visual


axes,


within


their visual


fields.


In freshwater


fishes,


,,,1:,,S


L. -- 2 -- ~- S I .c- -- -..- .


4


Li


L --. A L A


L- ->









within


the region


of its visual


field


just


above


the water


surface


(Schwassmann


and Kruger,


1965b).


Fernald


(pers


comn.)


recently


found preliminary

African cichlid H


evidence


.aplochromis


some retinotectal

burtoni, a freshwat


heterogeneity

er teleost in


in the


which


visual


signallin


and recognition


play


an important


role


in the animal's


lifestyle


(Fernald,


1977).


contrast


to the freshwater


fishes


most


not all)


marine


fishes


whose


retinotectal


patterns


have


been


mapped


have


shown


varying


types


of retinotectal


nonlinearities,


mostly


consistin


of temporal


retinal


areas


with


magnified


tectal


representations


(Schwassmann,


1975).


This


eterogeneity


reaches


a peak


in the serranid


fishes


(sea ba


sses


one of


the few teleost


families with


foveate


retinas


(Walls,


1942).


The fovea,


a morphological


depression


in certain


vertebrate


retinas


including


human,


an adaptation


associated


with


higher


visual


acuity.


In three


species


of the Pacific


serranid


Paralabrax,


Schwassmann


(1968)


found


a more


than


five-fold


retinotectal


magnification


of the


degree-wide


area


surrounding


the fishes'


visual


axes,


which


run from


their


rostral


visual


fields


to their


caudally


positioned


foveas.


Para-


labrax


a visually


orienting


fish,


with


adaptations


for specialized


spatial


acuity


that


include


binocular


convergence


of both


eyes


fixate


visual


targets,


keyhole-shaped


pupils


and sighting


grooves


snout,


all of which


act in


concert


with


primary


visual


axis


intersecting


the fovea


(Schwassmann,


1968).


Nonlinearities


in the retinotectal


projection


pattern


a species,


- I I- S I I S S


1


I


**


I


- 1









fibers


to the midbrain


roof


it follows


that


heterogeneities


in the


pro-


section


pattern might


reflect


topographic


differences


in ganglion


cell


distribution


across


the retina.


This


ganglion


cell


topography,


turn,


could


reasonably


be expected


to reflect


spatial


heterogeneities


in the


original


light-sensing mosaic,


the photoreceptor


layer.


Thus,


topogra-


phic


surveys


of ganglion and


photoreceptor


cell


densities,


when


con-


ducted


neural


in conjunction with


bases


retinotectal


for specializations


mapping,


in spatial


can further


vision


define


a species.


Retinal


Topo


raphy


As the light-detecting


tissue


in which


the photic


image


trans-


duced


ordered


into


electrochemical


meshwork


signals


neurons


glia.


the retina


The neural


is composed


elements


a highly


include


primary


photoreceptive


cells


(rods


and cones),


interneurons


such


as the


bipolar


and horizontal


cells,


and the tertiary-level


ganglion


cells


whose


axons


exit


the retina


via the optic


nerve.


Since


these


various


cells


are responsible


for the detection


and initial


encoding


of the


retinal


image,


it follows


that


the spatial


organization


and activity


patterns


of retinal


neurons


impose


certain


limits


on virtually


every


aspect


of the performance


sensitivity,


temporal


of the overall


response,


visual


color vision,


system,


and spatia


including visual

1 resolution.


According


to Hughes


(1977)


as far back as


1604


Johannes


Kepler


acknowledged


that


human


peripheral


vision


is less


acute


than


central


vision,


and Kepler


cited


variations


in retinal


structure


along with


poor


peripheral


optics


as being


responsible


for the differences.


The role









not only


across


the human


retina,


but also


between


species.


This


laid


roundwork


comparative


studies


of vertebrate


retinal


adaptations


for high


resolution


vision,


later


punctuated


the works


of Franz


(1934),


Walls


(1942),


Rochon-Duvigneaud


(1943)


Polyak


(1957),


and Duke-Elder


(1958)


Quantitative


studies


of retinal


topography,


particularly


of the


ganglion


cell and photoreceptor


layers,


have


recently


come


into


promin-


ence,


but as early


as 1935


0sterberg mapped


the densities


rod and


cone


cells


across the


human


retina.


Later,


Buren


(1963)


provided


quantititative


description


anglion


cell


topog


raphy


in human


retina.


More


recently,


Dawson


and Maida


(1984)


compared


0sterberg


s and Van


Buren's


data


with


visual


acuity


data and


found


very


high


correlations


between


vs. acuity


exceptions


cone

(0.99)


(Hughes


and ganglion


cones


1977),


cell


densities


vs. acuity


the distributional


(r=0.99),


Althou


patterns


ganglion

gh there


cones


cells

are some


gan-


glion


cells


in vertebrates


retinas


in general


are strongly


correlated,


with


peak


densities


in both


indicating


regions


of specialization


spatial


vision


for a particular


species.


With


the application


of retinal


wholemounting


(flatmounting)


tech-


niques


(Stone,


1981)


, topographic


mapping


of the


cone


and ganglion


cell


populations


of primarily mammalian


retinas


has been


upgraded


from


prev-


ious


studies


utilizing


sectioning methods.


Actually,


wholemounts


were


first


used


Schultze


in 1866


to study


retinal


organization


a variety


of diurnal


and nocturnal


vertebrates,


and this


eventually


led to the


-~ ~~~~~ a ,


n 1


C r. 1


mt


C









chromatic,


high


acuity


low sensitivity,


"photopic"


vision


in bright


light,


whereas


the rods


mediate


achromatic,


low acuity,


high


sens


itivity,


"scotopic"


vision


in dim light.


Since


the work


of Schultze


in 1866


, studies


on the relationship


between


comparative


retinal


topography


and the visual


ecology


of animals


have


been


conducted


primarily


on mammals


over


past


two decades,


per-


haps


following


in the wake


of Walls'


(1942)


writings


emphasizing


evolutionary


adaptations


comparative


ocular


design.


most


com-


pelling


overview


of this


subject


since


Walls


is that


of Hughes


(1977)


which,


although


written


as a review


of the mammalian


literature,


occa-


sionally


dealt with


nonmammalian


vertebrates


as well.


More


recently,


Stone


(1983)


has also


addressed


the relationship


between


retinal


topo-


graphy


and comparative


visual


function,


in his


review


anglion


cell


classification


in mammals.


Both


of these


reviews


emphasis


differen-


tial


organization


of the retina


along


taxonomic


lines,


although


Stone


position


is less


adaptationist


and phylogenetically


flexible


than


that


of Hughes.


For fishes,


analogous


studies


have


been


comparatively


few.


Kahmann


(1934,


Tamura


1936)


(1957)


compiled


Tamura


a list


and Wisby


of foveate


(1963)


species


of marine


and Fernald


(1983)


fishes,


demonstrated


a correlation


between


retinal


regions


of higher


cone


and ganglion


cell


density with


the axis


of accommodative


lens movements


in teleosts.


This


primary


line


of sight


the visual


axis


was


hypothesized


spec-


the fish


s zone


optimum


resolving


power,


loosely


referred


n cvrhnnhvqi rml


mpthnd-c


to mepasre


cornntrast


thresh-


1


tfn El a


a r.m tev


TTcni "no









roughly mapped


from sectioned


goldfish


retina.


Results


consistent


with


this


were


later


obtained


with


goldfish


using


psychophysically measured


visual


discrimination


tasks


as well


(Penzlin


and Stubbe,


1977).


Topographic maps


cone


density


in three


species


the cichlid


Haplochromis were


produced


by van


der Meer


and Anker


(1984)


and they


discussed


the ecological


significance


of retinal


structure


in these


fishes,


but only with


respect


to the relationship


between


cone


size


ambient


light


levels


during


feeding


activity.


On the other


hand,


cone


density


and ecologically


relevant


resolving


power


have


been


explored


the bluegill


sunfish


(Lepomis


macrochirus)


across


ontogenetic


lines


(Hairston


et al.,


1982).


In these


fish,


continuous


retinal


growth


throughout


life


produces


a decreasing


intercone


spacing,


measured


minutes


of visual


capture


small


angle,

r prey


such

than


that


larger


juvenile


fish


fish.


are able


This


to discriminate


study was


significant


not only


for its ramifications


for the visual


ecology


feeding


fishes,


but also


for its


support


of the Helmholtzian


view


that


retinal


resolving


power,


and ultimately


visual


acuity,


are fundamentally


limited


spacing


between


adjacent


cones


in the retina


(von


Helmholtz,


1924).


Finally,


the most


thorough


treatment


of the linkage


between


cone


density


ganglion


cell


density,


retinotectal


projection


pattern,


visual


axis


in fishes


has been


through


the studies


of Schwassmann


(1968)


and Schwassmann


and Meyer


(1971)


on the marine


serranid


Paralabrax.


Schwassmann


(1968)


mapped


the double


cone


and ganglion


cell


distribu-


1 S *- I 1- --


*t 1. .


1


--7-


1


*









subsequent


confirmation


Schwassmann


and Meyer


(1971)


that


the major


accommodative


axis


in these


fish


coincides


with


their


specialized


reti-


nal and retinotectal


zones,


the situation


for foveate


teleosts


was


cla-


rified:


lens


movement,


retinal


topography,


and retinotectal


projection


are matched


in register with


each


other


to afford


enhanced


spatial


vision


along


the fish


s visual


axis


which


a behaviorally


and ecolo-


gically


relevant


line


of sight


for the species.


Topography


of Spatial


Vision


in Sharks


Tectal


Anatomy


and Physiology


In most


chondrichthyan


fishes,


the mesencephalic


tectum is well


developed


into


two bilateral


lobes


(Fig.


each


of which


enclose


expan-


sions


of the ventricular


cavity


(Smeets


et al.,


1983).


shark


tectum


is the primary


termination


site


for retinofugal


fibers.


In all sharks


studied t

tectum is


o date, th

complete,


Le decussation


with


of these


the exception


fibers


to the contralateral


of extremely weak


ipsilateral


projections


to hypothalamic,


thalamic,


and pretectal


sites.


These


investigated


shark


species


include


the squalomorph


Squalus


acanthias


(Northcutt,


1979)


and the galeomorphs


Ginglymostoma


cirratum


(Ebbesson


and Ramsey,


1968;


Luiten,


1981)


Hemiscyllium


(=Chiloscyllium)


plagiosum


(Jen


et al.,


1983),


Scyliorhinus


canicula


(Smeets


1981;


Reperant


et al.,


1986),


Galeocerdo


cuvieri


(Ebbesson


and Ramsey,


1968),


and Negaprion


brevirostris,


the lemon


shark


(Graeber


and Ebbesson,


1972a).


The major-




















Left


Optic


Tectum


.Telencephalon


*Right Optic
Tectum






Cerebellum


Optic


Nerve


Fig.


Schematic diagram of position of mesencephalic optic tecta in


cranium of juvenile lemon shark.


Ganglion cell fibers from


retina of left eye cross over completely underneath the brain
and project to the roof of the right optic tectum.









Because


the elasmobranch


tectum


not nearly


as distinctly


layered


as the intensively


studied


tectum


of teleost


fishes


anatomical


nomen-


clature


of the shark


tectum has


been


a subject


of confusion.


Beginning


with


Houser


(1901)


the elasmobranch


tectum has


been


subdivided


dif-


ferent


authors


into


between


and 13 multiple


laminae


zones.


four-stratum nomenclature


of Ariens-Kappers


et al.


(1936)


dominated


literature


for about


years,


but with more


attention


focused


on elas-


mobranch


neuroanatomy


in the 1970's,


at least


six new nomenclatures


appeared.


Recent


attempts


to consolidate


these


descriptions


have


been


published


Northcutt


(1978)


and Reperant


et al.


(1986).


What


does


appear


to be generally


agreed


upon


however,


is the basic


pattern


of optic


fiber pathway


and termination within


the elasmobranch


tectum.


Contrary


to the situation


with


teleosts,


the optic


fibers


sharks


not enter


tectum


from


its external


surface,


rather


project within


the deep


tectum and


turn


outward


towards


the dorsal


zone,


terminating


in the superficial


layers


(Northcutt,


1977


Reperant


et al.,


1986).


Thus,


although


differentiation


of specific


tectal


layers


poor


compared with


that


teleosts,


seems


clear


that


the site


of optic


fiber


termination


is strictly


limited


to the


outer


tectal


zone


in all


sharks


studied,


as well


as in the skates


Raja


clavata


and R.


eglanteria


(Witkovsky


et al.,


1980).


For this


reason,


and for the sake


of simplicity within


scope


my study,


am choosing


to adopt


the streamlined


nomenclature


of North-


cutt


(1977,


1978),


recognized


three


tectal


zones


superficial,


a .


Ii





i


m


a









1978).

gical


It is in this


mapping


superficial


of the retinotectal


zone,


therefore,


projection


pattern


that


electrophysiolo-


should


be conducted.


Fortunately,


the superficial


zone


is also


the tectal


region


most


acces-


sible


for microelectrode


recording


Studies


on the


physiology


the elasmobranch


tectum


, especially


single


unit


or multi-unit


"hash"


responses


evoked


photic


stimuli,


have


been


few and far between.


Gilbert


et al.


(1964)


first


demonstrated


that


evoked


potentials


to light


flashes


could


be recorded


at single


sites


in the tectal


of small


lemon


sharks


(Negaprion


brevirostris)


and bonnet-


head


sharks


(Sphyrna


tiburo),


and similar


studies


on other


elasmobranch


cies were


reported


Karamian


et al.


(1966)


and Veselkin


and Kova


(1973).


Platt


et al.


(1974)


described


averaged


potentials


evoked


photic


stimulation


in Raja,


Torpedo,


Trygon,


cyliorhinus,


and Mus-


telus,


such


potentials


being


primarily


large


slow waves with


little


spa-


tial


localization


within


the mesencephalon.


Bullock


(1984)


mentioned


having


recorded


presumably


similar


responses


in the


tecta


of the sharks


Carcharhinus


melanopterus


and Negaprion


acutidens


(the


lemon


shark


of the


Indo-Pacific)


and the


rays


Rhinobatos


, Platyrhinoidis,


and Potamotrygon.


Bullock


(1984)


summarized


the body


of literature


available


on tec-


tal physiology


attention


in elasmobranchs,


to the omission


and,


reports


as previously

on elasmobranch


indicated,


draw


retinotectal


topo-


graphy.


In doing


so, he repeatedly


called


for new


studies


on single-


unit


characteristics


and the topographical


representation


of the visual


field


on the


tectum.









elasmobranchs


as a rule


possess


all-rod


retinas.


Although


there


were


several


early


reports


of duplex


retinas


in elasmobranchs


(see


Gruber


and Cohen,


1978)


the cone-free


viewpoint was


expressed


in the influen-


tial


writing


of Schultze


(1866)


Verrier


(1930)


Walls


(194


Rochon-


Duvigneaud


result


was


(1943),


a general


and Duke-Elder


(1958)


characterization


and this view prevailed.


of sharks


as functioning visually


with


pure


rod retinas,


and thus


being


"poorly


adapted


for distinguishing


the details


and color


an object"


but "well


equipped


S. for differ-


entiating


an object,


particularly


a moving


one,


from


its background"


(Gilbert


1963:


321).


The idea


that


sharks


were


strictly


cial-


ized


for nocturnal


vision


with


poor


resolving


power


and achromatic


vision


was


under photopi

perpetuated


conditions,


as recently


became


as 1975


ingrained

(Wolken, 1


in the literature


975).


However,


since


1963,


when


Gruber


et al. published


conclusive


evi-


dence


cones


in the retina of


the lemon


shark


(Negaprion


brevirostris),


many


researchers


have


confirmed


presence


cones


in a wide


variety


of elasmobranchs,


comprising


at least


24 species


in 10 families


summar-


ized


Gruber


Cohen,


1978).


The only


exception


that


has held


appears


to be the all-rod


retina


of the skate


Raja,


which


has thus


served


as a model


for vertebrate


visual


studies


in which


cone


function


is absent


(e.g.


Dowling


and Ripps,


1970).


Although


no researcher


to date


has systematically


plotted


cone


dis-


tribution


across


a shark


retina,


rod-to-cone


ratios


are often


reported


from sectioned


tissue.


Gruber


et al.


(1963)


reported


a rod-:cone


ratio


.


1. --









(Gruber


et al


., 1975)


a high


of 100:1


in the smooth


dogfish,


Mus-


telus canis


Stell


and Witkovsky,


1973b).


Only


five


reports,


however,


have


present


evidence


any topographical


heterogeneity


cone


dens-


across


retinas


of sharks.


In four


out of


the five


lower


rod:


cone


ratios


(and


thus


higher


cone


concentrations)


were


found


in central


retina


peripheral


retina,


as follows:


7:1 vs. 12:1


in Ginglymostoma


cirratum


(Hamasaki


and Gruber


, 1965);


vs.


13:1


also


in G.


cirratum


(Wang,


1968);


20:1


vs.


50:1


in Hemiscyllium


(=Chiloscyllium)


plagiosum


(derived


from


et al.,


1984)


and 4:1


vs.


10:1


in Carcharodon


car-


charias


(Gruber


et al.,


1975


Gruber


and Cohen,


1985)


Cohen


(1980)


mentioned


a rod:cone


ratio


5:1 in the "dorsal"


retina


of the lemon


shark,


as opposed


to the 12:1


count


reported


Gruber


et al.


(1963)


and Wang


(1968)


but Cohen


not further


specify


the retinal


location


this


higher


cone


concentration.


Because


all of these


studies


utilized


histological


sections


selec-


ted at random


from arbitrarily


divided


regions


of shark


retina,


very


little


functional


significance


can be inferred


from


these


data


on rod:


cone


ratios,


other


than


the implication


that


some


shark


species


have


regional


finding


retinal


a higher


specializations


cone


for spatial


concentration


vision.


in basically


Certainly


central


, the


vs.


peri-


pheral"


retina


not surprising


in a duplex


retina.


On the other


hand,


somewhat


better


topographic


information


is avail-


able


on gang


lion


cell


distribution


in sharks


, if only


in three


species.


Franz


(1931)


described


eyes


of Scyllium


(=Scyliorhinus)


canicula


Ml ~trol 11i


1 mPvi $


(=-m IjjI=,tII,


%s hnrth


rnn+tani f in


a linPer


nrsa


repntrlis









area.


From vertical


sections,


Franz


estimated


the ganglion


cell


concentration


in Mustelus


increases


from


800 cells/mm2


in the periphery


to 2500


cells/mm


inside


the band.


However,


Stell


Witkovsky


(1973a),


although


not disagreeing with


the relative


proportions


in total


cell


density


from outside


to inside


the band,


questioned


whether


these


cells


in Mustelus


were


all ganglion


cells,


suggesting


that


as many


as 90%


the cells


counted


Franz


could


have


been neuroglia.


The only


sharks


other


was reported


quantitative


Peterson


study o

and Rowe


f ganglion

(1980) fo


cell


topography


r the horn


shark,


Heterodontus


francisci.


Using


a wholemount


method


combined


with


cresyl


violet


staining,


Peterson


and Rowe


plotted


the density


of putative


gan-


glion


cells


at 1


mm intervals


across


five


horn


shark


retinas.


They


reported


that


the ganglion


cell


layer


of Heterodontus


is composed


pri-


marily


of cells


distribution


tha


averaging

t forms a


about


16 um in soma


horizontally


extended


diameter,


band


with


across


a peak

the retina


from


nasal


to temporal


periphery,


not unlike


the situation


in Scylio-


rhinus


and Mustelus


reported


Franz


(1931).


However,


the peak


density


of ganglion


cells


in Heterodontus


is 500 cells/mm


in the midline


of the


band,


and density


drops


off to between


50 and 100 cell


s/mm2


at the dor-


sal and ventral


periphery.


These various

and ganglion cell


reports or

populations


topographic


shark


heterogeneities


retinas


in the


are summarized


cone


in Table


Few of these


authors


addressed


the functional


importance


of the observed


retinal


heterogeneities,


Gruber


and Cohen


(1985)


speculated


that


-----------S


* 1 -.


1


*1-


2'


I'


__


__









Table


Previous


reports


of regional


specializations


cone


gan-


glion


cell


distributions


in shark


retinas


Specialization


Species


Source


Higher


cone


concentration


Ginglymostoma


cirratum


Hamasaki


in "central"


retina


Gruber


(1965)


Wan g


Hemi


scyllium


(=Chilo-


(1968)
et al.


scyllium)

Carcharodon


plagiosum

carcharias


(1984)
Gruber


Cohen


(1985)


Higher


cone


concentration


Negaprion


brevirostris


Cohen


(1980)


in "dorsal"


retina


Higher


gang


lion


concentration


horizontal


cell


along


meridian


Scyllium
rhinus)

Mustelus


=Scy


lio-


Franz


(1931)


canicula


laevis


Franz


(1931)


(=mustelus


Mustelus


canis


Stell
kovsk


and Wit-
y (1973a)


Heterodontus


francisci


Peterson


Rowe


(1980)









significance


of the horizontal


bands


of higher


ganglion


cell


density


reported


Franz


(1931)


in Scyliorhinus


and Mustelus was


later


dis-


cussed


Munk


(1970)


and Hughes


(1977);


the implications


of these


studies,


particularly


in relation


to the


retinal


topography


of the lemon


shark,


will


be taken


later


in the Discussion.


Studies


of Spatial


Visual


Performance


Although


an extensive


body


of research


on the visual


systems


sharks

exists


has emerged

on their ca


in the last


abilities


twenty years,


for handling


very


spatial


little


information


information.


Con-


trolled


experiments


to test


visual


discrimination


in sharks


began


in the


early


1960


s with


simple


instrumental


conditioning


techniques,


never


progressed


substantially


beyond


that


initial


stage.


difficulties


working with


documented


sharks


(Myrberg,


under


1976;


long-term,


Gruber


controlled


and Myrberg,


conditions


1977)


have


these


been


have


restricted whole-animal


studies


of shark


vision


to only


a few labora-


stories with


proper


facilities.


Unfortunately


visual


acuity


been


studied


in any


of these


laboratories.


Attempts


to train sharks


under


somewhat


controlled


conditions


locate


a target


for food


reward,


potentially


using visual


cues,


were


first


reported


Clark


(1959).


She was


able


to train


adult


lemon


sharks


to press


a white


cm square


target


to obtain


food,


and later


Clark


(1961,


1963)


also


trained


lemon


sharks


to visually


discriminate


between


a square vs.


a diamond,


and between a white


square


vs. one with


vertical


black and white


stripes.


But she did not


quantify


such


factors


not









basic


demonstration


that


sharks


could


learn


certain


visually mediated


tasks--which,


at the time,


was


nevertheless


newsworthy


The studies


that


followed


took a


similar


approach,


emphasizing


behavioral


traits


such


as acquisition


and retention


rather


than


limits of visual p

technique to train


performance.


juvenile


Wright


lemon


and Jackson


and bull


(1964)


(Carcharhinus


used Clark's

leucas) sharks


press


a white


cm square


target


for food,


but no discrimination


tasks


between


targets


were


involved.


With


similar


instrumental


condi-


tioning


techniques,


Aronson


et al.


(1967)


trained


juvenile


nurse


sharks


(Ginglymostoma


cirratum)


to discriminate


between


illuminated


vs. not


illuminated


x 11


cm targets.


The sharks


in this


study


learned


task


as effectively


and as quickly


as teleosts


(cichlids)


and mammals


(mice)

notion


trained


that


under


sharks


similar


were


conditions,


untrainable,


putting


unpredictable


to rest


the passe


experimental


subjects.


Using


aversive


conditioning with


electric


shock,


Tester


and Kato


(1966)


trained


juvenile


Pacific


reef


blacktip


(Carcharhinus


melanopterus)


gray


reef


menisorrah


= amblyrhyncho


sharks


to avoid


certain


visual


targets


and thereby


visually


discriminate


between


horizontal


vertical


white


rectangles


and between


a white


square


or white


circle


vs. a white


triangle.


attempt


was made


to find


a threshold


for vis-


ual discrimination


of these


shapes.


only


other


laboratory


to report


use of visual


discrimina-


tion


tasks


in shark


behavioral


studies


was


that


of Graeber


and his col-


leagues


(Graeber,


1978).


Experimenting with


juvenile


nurse


sharks


A- _-l2- 2_! ? .2 ._ -_ 2-


^-1- -I 1-


\T ---- ----





-- --


--- I-r-~


A


.m #' u A L- -


Ln









cm wide,


with


each


target


consisting


of three


black and


three


white


stripes.


sharks


were


forced


to choose


between


targets


from


a dis-


tance


no less


than


cm away.


By my


calculations,


this corresponds


a visual


angle


of 190


for a half-period


of the grating


(width


stripe


at minimum viewing


distance)


or less


than


0.03


cycles


per degree


c/deg).


This


is certainly


not a task


requiring high


visual


acuity


rela-


tive


to the capabilities


of other vertebrates.


However,


there


no way


to know


at what


distance


from


targets,


between


cm and 2


m away,


that


the sharks


discriminated


the stripe


pattern,


and Graeber


push


the sharks


to threshold.


Graeber's work


drawn more


attention


for his


use of this


dis-


crimination


paradigm


assess


visual


function


nurse


sharks


after


ablation


of brain


structures.


Graeber


and Ebbesson


(1972b)


and Graeber


et al.


(1973)


stated


that,


at least


one shark,


bilateral


tectal


abla-


tion


not prevent


the learning of


the visual


discrimination


task,


although


in the


the shark


postoperative


bumped


into


period.


the aquarium walls


researchers


and objects


suggested


that


earlier


recovery


of visual


function


been mediated


possibly


some


other


area


of the


CNS,


they


suspected


telencephalic


involvement,


which


they


later


investigated


(Graeber


et al.,


1978).


case


Graeber's


findings


raised


questions


about


the specific


role


of the shark


optic


tectum in


processing


spatial


visual


information.


have


previously


attempted


to estimate


the potential


resolving


power


the juvenile


lemon


shark


eye,


based


upon


retinal


magnification


rnl 'lla nct1 a


Frnrn n


a rknhmnr4 p


In-r t'flo -r rIIti in


rnio-h


one


not


-fa~ nrr


1QI nO"


i,4 T |


CI\/CI









grating


target


like


that


used


in Graeber's


tests,


was


a visual


angle


arc,


which


more


than


two orders


of magnitude


finer


than


conditions


of the Graeber


test.


This


perhaps


underscores


the need


more


stringent


testing


of visual


discrimination


in both


intact


ablated


sharks


before


we can draw


specific


conclusions


about


mesenceph-


alic


vs.


telencephalic


control


of visual


function.


Juvenile


Lemon


Shark


Biology


of the Lemon


Shark


From


the previous


are a long way


from


review


understand


of the literature, i

ing the fundamentals


t is evident

of spatial


that

vision


sharks


much


less


its role


in the physiological


ecolo


shark


species.


A comparative


wealth


of information


exists


on such


features


as absolute


sensitivity,


adaptation


dynamics,


ocular morphology,


and cellular


struc-


ture


and physiology


certain


elasmobranch


visual


systems


but the


spatial


organization


of those


systems--retinal


topography,


retinotectal


projection


patterns,


functional


acuity,


etc.


--has


so far eluded


fact,


discussions


shark


vision


almost


never mention


concept


true


visual


axis,


due to the lack


of information


in this


arena


the confusion


surrounding whether


or not sharks


can accommodate


(Sivak,


1978).


This


has in


turn


crippled


the effectiveness


of studies


to assess


the visual


performance


of sharks,


and has frustrated


efforts


to charac-


terize


the optics


and resolving


power


of the shark


(Hueter


and Gru-


' of


us.


I-








There


are approximately


300 species


of living


sharks


presently


des-


cribed

prion


(Compagno,

brevirostri


1977).


as a mode


choice of

1 species


the juvenile


to b


lemon


egin answering


shark

these


(Nega-

ques-


tions


was


based


on several


factors.


Most


important


of these


was that


extensive


optical,


anatomical,


electrophysiological,


and psychophysical


studies


of the visual system in


this animal


have


been


conducted


Gruber,


1967


, 1975;


Cohen


et al.,


1977;


Hueter,


1980),


and retinal


pro-


sections


and tectal


anatomy


have


been


previously


described


for this


species


(Graeber


and Ebbesson,


1972a).


In addition,


the juvenile


lemon


shark offers


certain


advantages


laboratory


study.


In Florida,


it is available


nearly year-round,


parti-


cularly


on the flats


of Florida


in south


Florida,


and compared


with


other


species


is relatively


transportable


from


the collecting


site


the laboratory.


Specific


recommendations


for maintaining


this


shark


captivity


for research


purposes


have


been worked


out (Gruber


and Keyes,


1981).


Furthermore,


the lemon


shark


among


the few shark species with


the ability


to pump water


over


the gills


and continue


to respire


while


resting,


which makes


more


amenable


to confinement


in the laboratory


than most


other


species.


Another


consideration


in choosing


the lemon


shark


is its phyletic


position


among


the selachians:


as a carcharhinid,


it is a member


of the


dominant


group


of galeomorph


sharks,


which


comprise


73% of all living


shark


species


(Compagno,


1977).


Thus,


the lemon


shark


not a bad


choice


as representative


the mainstream of


shark


evolution.


Further-


more,


as with


other


carcharhinids,


the lemon


shark


grows t


a large


size









cases


from


the international


Shark Attack


File


in which


species


shark


could


be identified,


six cases


cited


a lemon


shark


as the attacker


(Baldridge,


1974).


The lemon


shark


a common


inshore


species


ranging


in the western


Atlantic


from New


Jersey


to southern


Brazil


(Compagno,


1984)


and the


juveniles


are often


most


abundant


shark


in the shallow


coastal


waters


of Florida


and the Florida Keys


(Springer,


1950).


According


to Gruber


and Stout


(1983),


the lemon


shark


a slow-growing


late-


maturing,


long-lived


speci


reaching


sexual


maturity


no less


than


Mature


females


may reproduce


every


other year


(Clark


Schmidt


, 1965)


giving


birth


to an average


pups


after


a gestation


period


of about


(Gruber


and Stout,


1983).


The adult


sharks


less


abundant


in the shallower waters


of Florida


in the latter


months


of the


year


pringer,


1950)


presumably


because


the spring


early


summer


reproductive


season


over.


The juveniles


remain


in the


Florida


shallows


for at least


their


first


year,


in which


they


grow


approximately


cm from


their


birth


size


of about


cm total


length


(Gruber


and Stout,


1983).


The only


potential


predators


of the juveniles


are larger


sharks


and the


young


sharks


probably


continue


to reside


shallow waters


around


the Florida


Keys


until


they


reach


sexual


maturity,


when


they


assume


a somewhat


more


offshore


distribution.


On the flats


of Florida


Bay,


the juveniles


are active


predators


feeding


primarily


on fishes


some


crustaceans.


Springer


(1950)


served


juvenile


lemon


sharks


feeding


on mullet


(Mugil


spp.)


and Starck


, -, a r a x I e a 4 4 I


a


von


are





* 1 1 I


1


*)


r- *


1


rn\ / r


/n rt









crawfish


(Florida


spiny


lobster,


Panulirus


spp.)


crabs


"and,


curiously


lots


of turtle


grass"


in lemon


shark


stomachs.


The last


item


be related


to incidental


ingestion


of benthic


vegetation


during


sharks


' pursuit


crustacean


prey:


bonnethead


sharks


(Sphyrna


tiburo),


which


also


inhabit


the Florida


shallows


and feed


heavily


on small


spiny


lobster,


often


show


large


amounts


of turtle


grass


in their


stomach


contents


as well


Parsons,


pers.


comm.).


Adult


lemon


sharks


continue


to feed


on fishes


crustaceans,


with


some


evidence


of specialization


on stingrays


(Gruber


1981).


Among


prey


items


taken


lemon


sharks


throughout


its distribution,


Compagno


(1984)


lists


mainly


teleosts


(sea


catfishes,


mullet,


acks,


croakers


porcupine


fishes,


cowfishes)


followed


elasmobranchs


(guitarfish,


stingrays,


and eagle


rays),


crustaceans


(crabs,


crawfish


, barnacles,


and amphipods),

It is clear


molluscs

that th


(conchs)


e juvenile


and

lemo


occasional

n sharks a


sea birds.


re patrolling over


flats


very


shallow water,


seeking


their


food


on or just


over


bottom.


Although


the adults


move


into


deeper water,


they


remain


primar-


benthic


feeders,


and they


retain


the bottom-resting


trait.


Even


where


lemon


sharks


are common,


they


are not normally


seen


at the surface


among


other


surface-swimming


sharks


(Springer,


1950).


Contrary


to the


common


generalization,


neither


juvenile


nor adult


lemon


sharks


appear


to be restricted


to nocturnal


periods


of activity.


Starck


(1968)


reported


that


active


feeding of


the juveniles


in Florida


occurs


both night


and day,


especially


at low tide,


and Gruber


(1982)


__








rising


setting


sun (Gruber,


1982).


This


indicates


that


the lemon


shark


is crepuscular


like


some


other


reef


shark


species,


but continues


to be equally


active


throughout


the day


and night


(Gruber,


1981).


Visual


System


Features


The habits


of the lemon


shark


would


seem


to call


for a versatile,


duplex visual


system


that


can


adapt


to both


photopic


and scotopic


con-


editions.


In fact,


behavioral


and electrophysiological


studies


of juven-


ile lemon


sharks


in the laboratory


have


revealed


extensive


evidence


duplex


visual


function,


from


the standpoints


of spectral


sensitivity


dark


adaptation


dynamics,


color


vision,


and other


features


(reviewed


Gruber,


1975).


eyes


of the juvenile


lemon


shark


are mounted


laterally


on its


relatively


broad


head,


with


only


a small


degree


of binocular


overlap


between


two monocular


visual


fields.


But as the shark


swims


side-to-side


sweep


motion


of the horizon


of the head

tal visual


gives

field.


the animal

Small, co


an essentially


mpensatory


3600


eye movements,


similar


to those


observed


Harris


(1965)


in the spiny


dogfish


(Squalus


acanthias)


appear


to adjust


for turning


of the head


during


swimming,


most


likely


to maintain


some


stability


of the retinal


image.


Externally


visible


features


of the


include


an active


vertical-


slit


pupil,


which


contracts


a thin


slit


under


photopic


conditions


(full


light


scotopic


adaptation),


conditions


(full


and dilates


dark


a near-circular


adaptation).


aperture


The nictitatin


under


membrane,


a feature


that


the lemon


shark


shares


with


other


carcharhinids









Intraocular


organization


of the lemon


shark


eye is fairly


typical


of vertebrate


layer,


eyes.


The sclera


and the nutritive


choroid


is supported

contains an


a thick


occlusible


cartilagenous

tapetum lucidum,


whose


reflective


plates


are actively


exposed


under


scotopic


conditions.


The nearly


in the


spherical,


eye (Hueter,


crystalline


1980).


lens


lens


is the sole


is supported


refractive


a dorsal


element


suspensory


ligament


and the ventral


pseudocampanule,


said


to function


as a protrac-


lentis


muscle


in lens


accommodation


(Walls,


1942).


However,


efforts


to confirm


this mechanism in


the lemon


shark


have


so far failed


(Sivak,


1974;


Hueter,


1980).


The retina


not


vascularized


and contains


no obvious


landmarks


other


which


nerve


than


marks


from


the optic


the point


retina


disc


(corresponding


of exit


to brain.


to the shark's


of the ganglion


There


cell


no fovea,


blind


fibers


spot),


the optic


no specialized


area


has been


previously


described


in this


species.


In histologi


cal sections


cones


are readily


differentiated


from


rods


on the basis


of size,


shape,


and staining


characteristics


(Gruber


and Cohen


, 1978).


Rods


are much


longer


than


cones,


with


a combined


outer


less


and inner


than


segment


length


for the rods


of 32 pm and


in fixed,


a nonvarying


dehydrated


tissue.


diameter


Cones


are


19 1dm long


and have


inner


segments


approximately


pm in


diameter


histological


sections.


cone


outer


segment


is tapered,


with a-


mid-length


diameter


of 2.5


The tips


of the


cone


outer


segments


reach


to the level


of the rod ellipsoids,


which


are smaller


than


-~~~~~~~~~n 4af i -a


^ Af- t\


t


. 1*


I


I









cone


density

(1968)


across

stated


the whole


that


retina.


the rod:cone


Both

ratio


Gruber


young


et al. (19

and adult


6


3) and

lemon


sharks


appeared


to be


a uniform


12:1.


However,


as previously mentioned,


Cohen


(1980)


indicated


finding


some


heterogeneity


this


ratio


with


counts


reaching


in so-called


"dorsal"


retina.


In the absence


thorough


topographic


surveys


of either


cone


or ganglion


cell


density


and without


evidence


an axis


of lens


accommodatory


movement,


primary


visual


axis


has been


described


for the lemon


shark.


Cohen

as falling


(1980)

into t


described


:wo classes.


the ganglion

The first


cells

type a


of the lemon


re small,


shark


spherical


retina

cells


measuring


8-10


in diameter


and containing


irregularly


shaped,


eccen-


trically


located


nuclei.


The second


type


are giant


ganglion


cells,


measuring


20-30


in diameter


and containing


spherical


nuclei


with


distinct


nucleoli.


Contrary


to most


other


vertebrates,


whose


ganglion


cell


axons


are nonmyelinated


until


they


exit


enter


the optic


nerve,


these


axons


in elasmobranchs,


including


the lemon


shark


myelinated


within


the retina


throughout


nerve


fiber


layer


(Cohen,


1980).


Extracellular


recordings


made


from


ganglion


cells


of lemon


shark


retina


have


revealed


three


types


of units


based


upon


response


to the


onset


and offset


a light


stimulus


(Cohen


and Gruber,


1985).


In these


recordings,


OFF units--those


cells


responding


only


to the offset


light--are


most


numerous


, comprising


63% of units


classified


Cohen


and Gruber's


study.


ON units,


responding


only


to the


onset


light,


made


18% of the units


and ON-OFF


units,


responding


to both


r 4CT/ a 1 I


are


I


~1


1


t


rF r I I


I


I -









visual


angle


of between


5.2 and 8.80


given


a retinal


magnification


fac-


tor of 0


.164


mm/O


from


the lemon


shark


schematic


eye (Hueter


and Gruber,


1982).


Graeber

the pathways


and Ebbesso

of retinal


(1972a)


projection


used


degeneration


and tectal


anatomy


techniques


to follow


in the lemon


shark.


They


reported


that


the optic


fibers


undergo


complete


decussation,


with


the exception


a very


small


number


of fibers


terminating


ipsilaterally


in the hypothalamus.


As the optic


tract


enters


the thalamus,


great


maj ority


of fibers


turn


dorsolaterally


and remain


in a compact


lateral


tract,


eventually


terminating


in the contralateral


optic


tectum.


few remaining


scattered


fibers,


according


to Graeber


and Ebbesson,


pear


to ascend


directly


to the dorsal


tectum.


All entering


optic


fibers


penetrate


the superficial


tectal


zone


(stratum


griseum


et fibrosum superficial


of Graeber


and Ebbesson)


from


underneath,


and terminate


in this


zone


out to the periphery


of the


tec-


turn.

optic


Compared

tectum,


with


the highly


the lemon


shark


differentiated


tectum


layering of


not as distinctly


the teleost

laminated,


with

Some


the densest

recognizable


termination

lamination


of fibers

of tectal


in the internal


neurons,


superficial


especially


zone.


in the lower


periventricular


zone,


has been


briefly


described


in the lemon shark


Klatzo


(1967).


However,


the comparatively


diffuse


nature


tectal


organization in


the lemon


shark


led Graeber


Ebbesson


(197


question


the role


of the tectum in shark


vision,


and spurred


their


studies


of tectally


ablated


nurse


sharks.


Precise


electrophysiological


-~~ 4


*


1 1


1


1












Objectives


of the Study


primary


purpose


of this


study was


to determine


if sharks,


exemplified


here


the juvenile


lemon


shark,


have


within


their


visual


system any


topographically


nonlinear


representations


of their


visual


field


that


would


indicate


a visual


axis


or axes


for specialized


spatial


vision.

of this


Should s

nonlinear


uch


differentiation


sampling


strategy


exist,


the functional


in the visual


ecolo


significance

of the lemon


shark


requires


attention.


To this


end,


sought


the following major


objectives:


map the precise


retinotectal


projection


of the visual


field


of the lemon


shark


onto


its optic


tectum


utilizing


electrophysiological


techniques,


and thereby


produce


the first


of the retinotectal


pro-


section


a chondrichthyan


species;


the relative


distributions


cone


and ganglion


cell


density


across


the lemon


shark


retina


as indicators


of topographic


heterogeneity


in retinal


organization,


in order


to provide


an anatomical


basis


for the spatial


pattern


of the retinotectal


projection;


assess


the implications


of such


topographic


organization


of spatial


vision


for the visual


ecology


of the lemon


shark,


taking


into


consideration


the demands


of habitat


and lifestyle


on the visual


system


of this


shark


species.















MATERIALS


AND METHODS


Animal


Collection


and Maintenance


Collection


Data


were


collected


from a


total


of 46 juvenile


lemon


sharks,


proximately


40% males


and 60% females,


during


course


of the study.


Mean


size


the sharks was


cm total


length


cm fork


length


(FL),


56 cm precaudal


length


(PCL)


and 2.1 kg


mass,


with


ranges


63-90


cm TL,


52-79


cm FL


, 49-71


cm PCL,


and 1.2-4.9


This


size


range


corresponds


to an age


of less


than


two


years


after


birth


(Gruber


Stout,


1983).


animals


were


collected


in the Florida Keys,


with


the assistance


of personnel


of the Sea World Marine


Science


and Conservation


Center


Long


Key,


Florida,


in an


area


of Florida


approximately


10 km NNE


the Long Key


facility.


This


area


is a major pupping


nursery


ground


for the lemon


shark,


and the juveniles


are relatively


abundant


over


turtle


grass


flats


and around


mangrove


islands


there


throughout


most


of the year.


Only


during


the three


months


between mid-January


and mid-


April


are the juvenile


sharks


somewhat


difficult


to find


in this


region,


until


the newborn


pups


begin


to appear


in spring.









flats,


chumming with


a mesh


containing


cut fish


parts


and blood.


Sharks

20-30


usually

min after


appeared


in the chum


chumming was


begun.


line

When


swimming

a shark


toward


the boat within


was within


m away


from


the boat,


a circular


hoop


of fencing wire,


measuring


1.5 m in dia-


meter


and 0.6 m in height,


was


thrown


over


the freely


swimming


shark.


The trapped


shark


was


then


dip-netted


from


inside


the hoop


a live-


well


in the boat,


With


these


in which


methods


was transported


the sharks


back


experienced


to the Long


a minimum


facil-


trauma


physiological


stress,


much


less


than


would


be produced


with


hook-and-line


or cast-net


capture


techniques.


At Long


Key,


the sharks


were


allowed


to adjust


to captivity


55,000


2-3 days


aquarium


containing


in captivity,


filtered,


the sharks


were


recirculating seawater.

prophylactically treated


After

with


addition


of 2.0 ppm dimethyl


(2,2,2-trichloro-l-hydroxyethyl)


phosphanate


(Dylox)


to the aquarium seawater.


This


chemical,


which


degrades


in sea-


water


after


24-36


controls


skin


infections


of Vibrio


bacteria


monogenetic


trematodes


(Dermopitherius


both


of which


can cause


major


problems


for lemon


sharks


kept


in captivity


(Gruber


and Keyes,


1981).


Following


a recovery


after


prophylaxis


several


days,


sharks


began


feeding


in captivity,


and were


ready


for overland


transport


to the laboratory within


a week


after


being


collected.


Transport


and Maintenance


Systems


to ten juvenile


sharks


at a time were


transported


live


unan-


esthetized


from Long


to the Whitney


Marine


Laboratory


in St. Augus-


sp.),









with


a 100%


constant


bubbler


seawater


filtration,


both


driven


a 12 V automobile


battery.


Whitney


Lab box


had an 02


bubbler


12 V


seawater


pump


to recirculate


the tank


water without


filtration.


both


cases,


all animals


survived


the 8


transport


and began


feeding


again


in captivity within


several


days.


At the Whitney


Laboratory,


the sharks


were maintained


several


months


a shallow,


semi-natural


seawater


pond


averaging


about


Im in


depth,


in an


enclosed


area


of the pond


approximately


m long


m wide.


This


pond


was


part


of the Laboratory's


open,


flow-through


seawater


system.


seawater


in the pond


was


sand-filtered


Atlantic


ocean


water,


34-35


ppt salinity


and 23C


average


temperature.


The water


temperature


ranged


from a


high


of 280C


in August


a low of 14C


February,


but this


seem


to affect


the survival


of the captive


sharks


or the


success


of the mapping


experiments.


The only


noticeable


effect


to temperature was


slug


ishness


very


low food


intake


the sharks


during


the coldest


periods.


The pond


was


not shad


and received


a natural


daily


cycle


sun-


light.


Underwater vis


ibility


in the pond


was


typically


about


The bottom


consisted


coarse mud


and sand


covered


with


algae,


free


of obstructions.


Other


fishes were


excluded


from


the shark


enclo-


sure


ensure


that


food


intended


for the sharks was


not taken


other


pond


residents.


During


course


of the study,


no sharks


died


in captivity


the shark


pond,


no animals


exhibited


external


signs


of serious


* I- s-


3


. . ---- I -- -


was


f


-_ -- -- _- _-- --1


1_ __ *


_ *


I _


I _


_ J .1 _


*









Food


and Feeding


fed the juvenile


lemon


sharks


a variety


of fresh


and frozen,


thawed


food


items.


Following


Springer's


(1950)


observation


that


these


sharks


feed


on mullet


in Florida


Bay,


prepared


cut pieces


of fresh


striped


mullet


(Mugil


cephalus)


including


bone,


skin,


and fin


parts


as a main


food


item


for the lemon


sharks.


Mullet


is a relatively


oily


fish,


and in


a small,


closed


aquarium system


not a practical


food


item


due to its introduction


oily


contaminants


which


produce


a sur-


face


scum


in the


aquarium.


a large,


open


system,


however,


this


not a problem,


and the juvenile


lemon


sharks


readily


fed on the


mullet.


In addition


to mullet,


the following


teleosts


were


also


fed to the


sharks


as fresh,


cut


pieces


in varying


amounts:


bluefish


(Pomatomus


sal-


tatrix)


killifish


(Fundulus


spp),


crevalle


jack


(Caranx


ladyfish


(Elops


saurus).


The only


fish


that


was tried


as a fresh


food


item and


was rejected


the sharks


was hardhead


catfish


(Ariopsis


felis),


delivered


as filleted


pieces


with


skin


a curious


finding


light


of Compagno


s (1984)


inclusion


sea catfish


in the diets


lemon


sharks.


When

of frozen,


fresh f

thawed


ish


fish


was not available


or invertebrates


in sufficient


were


amounts


fed to the sharia


, portions

ks. Among


the frozen


fishes


in decreasing


the following


order


usage:


species


striped


were


mullet


delivered

Atlantic


as food

thread


items

herring


(Opis thonema


oglina)


Atlantic


mackerel


(Scomber


scombrus),


Atlantic


menhaden


SAL


Irevoortia


tvrannus) -


and Atlantic


bumper


(Chloroscombrus


hippos)


J









fish


food


items.


In addition,


to guard


against


vitamin


deficiencies


particularly


plement


in the frozen


(formulated


foods,


for humans)


added


a standard


to the sharks'


food


multivitamin

at the dosage


sup-


one


tablet


shark


(about


40 mg


per kg


body mass)


each


week.


Fresh


or frozen


food


was delivered


in 20-50


pieces


on the fol-


lowing

sion,


schedule:

one day of


one day


: light


of heavy


feeding


feeding to satiation

one session, one day


a single


of fasting,


ses-

then


repeat.


With


this


schedule,


which


is comparable


not identical


the 3-


venile


to 4-day


lemon


feeding


sharks


les observed


in captivity,


Longval


animals'


et al.


interest


(1982)


in food


in ju-


remained


high.


Their maximum weekly


consumption


conformed


approximately


to the


figure


of about


25% body mass


of wet weight


food


reported


for young


lemon


sharks


in captivity


by Gruber


(1980).


Preparation


for Experiments


Sharks


selected


for experimentation were


transferred


from


the pond


an outdoor,


3800


seawater


tank


prior


to the day


of each


experiment,


greater


ease


of handling


on the day


of the experiment.


Sharks


were


removed


from


the pond


confining


them


a shallow


shoreline with


beach


seine,


and then


dip-netting


them


from


the pond


to the


concrete


tank.


Inside


dimensions


the rectangular


tank were


2.6 m long


2.0 m wide,


with


a water


depth


of about


0.8 m.


corners were


smoothed


out with


plywood


boards


to promote


circular water


flow


and guide


sharks'


swimming movements


away


from


the walls.


The tank seawater was


part


same


open,


flow-through


system


supplying


the pond.


Water









a naturally


timed


cycle


of daytime


and nighttime


illumination.


con-


tinued


to feed


animals


in the tank


along


the schedule


previously


des-


cribed,


prior


to their


use in


experiments.


To minimize


amount


of bleeding


during


pre-experimental


surgery


promoting


rapid


clotting


capabilities,


experimental


animals


were


given


a vitamin K


supplement


8-10


hr prior


surgery.


administered


an intramuscular


section


of Aquamephyton


(Merck


Sharp


Dohme),


aqueous


colloidal


solution


of vitamin Ki,


a dosage


of 2


mg per kg


body


mass


(about


mg per shark).


This


treatment


significantly


creased


success


of surgical


procedures


and electrophysiological


periments


on the following


day.


Retinotectal Mapping


Surgical


Experiments


Preparation


Experimental


animals


were


transferred


from


the outdoor


shark


tank


to the laboratory


conducted.


where


shark


was


preparatory


anesthetized


surgery


with


and vision


a knock-down


experiments


dose


were


of 1:1000


seawater)


of ethyl


m-aminobenzoate


igma) ,


also


called


tricaine


methanesulfonate


(MS-22


Sandoz


or Finquel,


Ayerst),


or ethyl


p-amino-


benzoate,


also


called


benzocaine


igma)


delivered


via a seawater


sprayer


directed


through


the gills


(Gilbert


and Wood,


1957).


When


shark


ceased


to strugg


le and slowed


all ventilatory movements,


it was


transferred


a surgical


arena,


where


was placed


on a regulated


flow


ex-









activity,


with


an intramuscular


injection


of 1.5


curare


body


mass


delivered


a solution


of 3 mg/ml


saline.


An approximately


cm patch


of skin


centered


over


the chondro-


cranium


e tween


eyes


was removed


using


a small


rotary


saw (Dremel)


and scalpel.


Bleeding


from vessels


in the


cut skin


was


stopped


using


electrocoagulator


(Sybron).


Cranial


cartilage


overlying


the midbrain


relatively


soft


in the juveniles,


was cut,


shaved,


and clipped


away


carefully with


scalpel


rongeurs,


exposing


the outer meninges.


Some


larger


blood


vessels


in this


tissue were


electrocoagulated


at the edge


of the chondrocranial


opening,


and the tissue


was


cut and moved


to the


side,


exposing


the cerebe


llum overlying


the midbrain.


The cerebellum


gently


retracted,


and cerebrospinal


fluid


(CSF)


was


aspirated


of the cranial


cavity.


Slow


seepage


of CSF kept


the midbrain


surfaces


moist


throughout


the right


tectum,


the experiment.


an adjustable


To fully


expose


retractor was


the dorsal


positioned


surface


to spread


apart


the telencephalon-tectum


boundary


and tectum-cerebellum


boundary.


Wedges


of Gelfoam


(Upjohn)


were


inserted


in spaces


around


and beneath


the tectum where


necessary,


and between


the dorsal


chondrocranium and


cut skin


to control


bleeding.


Extreme


care was


taken


to avoid


abra-


sion


of the tectal


surface


and damage


to the extensive


vascular


supply


of the brain


Mapping


and chondrocranium.


Chamber


The shark was


transferred


from


the surgical


arena,


on an acrylic


plastic


(Plexiglas)


adjustable


platform


lined with


polyurethane


foam,


was
















































Fig.


Experimental


sharks.
lyzed,
genated
cranial
mounted


Top,


chamber
rostral


anesthetized


seawater.


preparation


above.


the water-filled


for mapping


view;


shark


bottom,


receives


The shark's
is in air,


The left ey
hemisphere


with


retinotectal


left


side


a constant


projection


view.


supply


is underwater
the recording


positioned


representing


at the


the 1800


para-


oxy-


and the


electrode


center


left


visual


field.
visual
adapted


Photic
units a


stimuli


re mapped


from Schwassmann


are presented


and receptive


on the hemispheric
(1975).


surface.


fields


Apparatus









positioned


left


was


in the seawater-filled,


located at


clear


geometric


plastic


center


chamber,


so that


of a water-filled


hemi-


sphere


vertically mounted


to the side of the holding


tank.


The shark


restrained


on the platform


in fixed


position


with


a head


clamp,


hard


plastic mouth


tube,


and three


body


straps.


A regulated


flow


seawater


(approximately


1/min)


containing


:20,000


(0.05


tricaine/benzo-


caine


anesthetic was


continuously


delivered


through


the mouth


tube


over


the shark'


gills.


This


seawater


supply was


gravity-driven


from an


80 1


reservoir


in which


100%


was


bubbled


and the supply was


recirculated


throughout


the experiment,


using


a 12 V DC submersible


pump


to periodi-


cally


refill


reservoir.


seawater


level


in the holding


tank was


positioned


at least


above


the shark's


but below


the surgical


opening


into


the cranium,


using


an adjustable


standpipe.


With


this water


level


the surgical


pre-


paration was


kept


free of seawater without


interferin


with


the under-


water


optics


the left


eye,


so that


the water-filled


hemisphere


simu-


lated


the 1800


left


visual


field


of the shark


underwater.


A slight


gular


error


introduced


in the simulation


of the visual


field


the small


difference


in refractive


index


between water


(1.33


for the


distilled water


inside the hemi


sphere,


1.34


for the


seawater


in the


holding


tank)


and plastic


(1.48)


in the hemisphere


and tank wall.


Schwassmann and


Kruger


(1965a),


however,


have


calculated


that


this


error


is only


about


one degree


arc at the


extreme


periphery,


decreasing


zero


at the


center


of the visual


field,


and so I


ignored


it in


__ 1a


was


an-









center


meridians


from


of the visual


at 100

dorsal


field,


intervals.


to 900


ventral


with


left


and 90


field


lines


visual


rostral


parallel


ield w

to 90


as thus


caudal


to the


represented

on the


hemispheric


surface.


position


was carefully


checked


proper


alignment


of the


eye's optical


axis


with


the hemisphere


in the following ways.


First,


center


of the pupil


was positioned


at the intersection


cross


hairs


marking


the hemispheric


center


and drawn


on the


common


wall


tween


the holding


tank


and hemisphere.


(The


cross


hairs


were


interrup-


near


the ocular margin


so that


they


not occlude


the visual


field.)


This


alignment


was checked


to coincide


with


the crossing


the 0O


horizontal


and vertical


meridians


on the


outer


surface


of the


hemisphere.


Next


the pitch


of the head


was


adjusted


aligning


vertical


slit


pupil


with


the vertical


cross


hairs.


of the head


was also


adjusted


evenly


aligning


the corneal


surface


with


the hol-


ding


1-3 mm


tank wall


between


as viewed


dorsally,


the cornea and


wall.


with


When


an aqueous


ocular


separation


position


was


of about


acceptable,


the head


clamp


and body


restraints


were


tightened


to prevent


head


move-


ments.


No suturing


of the


was


necessary,


as the shark


showed


evidence


eye movements


during


experiment.


a final


check


on ocular


position,


located


the left


optic


disc


(optic


nerve


head)


by visualizing


it via


a direct


ophthalmoscope


placed


on the hemispheric


surface,


and mapped


the disc's


representative


loca-


tion


on the hemisphere


throu


the ophthalmoscope.


For the juvenile


* I S y


*^


-1


I I I .


-<


1


1 r\ / lo


^ ^y\u .









checked


to approximate


these


coordinates.


The disc


location


later


rechecked


periodically


throughout


experiment


to verify


that


ocular


position had


not changed.


Refractive


error


of the shark'


was not optically


corrected


for,


even


though


have


previously


shown


(Hueter,


1980)


that


juvenile


lemon sharks


are somewhat


hypermetropic


(farsighted)


averaging +2.76


diopters


of refractive


error.


This


ametropia would


distort


apparent


to-center


size


of receptive


distribution


fields,


of fields,


not affect


which


the relative


was the critical


mapping


center-


criter-


ion.


Even


calculations


using


the schematic


for the juvenile


lemon


shark


(Hueter,


1980)


show


that


ametropic


blur would


increase


apparent


size


of receptive


fields


a relatively


small


margin:


1% dis-


tortion


a field


wide,


2% distortion


a 50


field,


and 12% dis-


tortion


field.


Since


the ganglion


cell


receptive


fields mapped


in these


experiments were


generally


on the order


of 5-100


chose


ignore


refractive


error,


as have


other workers


Schwas smann


Kruger,


1965a;


Heric


and Kruger,


1965).


Photic


Stimuli


Stimuli


used


for mapping were


presented


on the


outer


surface


of the


hemi


sphere,


in front


a diffusely


reflecting


screen


located


approxi-


mately


cm from


the hemisphere.


tungsten


incandescent


light


source


illuminated

sphere with


screen,


a mean


producing


luminance


a white


background


of approximately


behind


1.6 logfL


(136


the hemi-


cd/m2),


as measured with


a SEI photometer.


remaining


peripheral


illumination


was


was









Throughout


experiment,


the pupillary


aperture


remained


a size


(about


18 mm 2)


and configuration


(slightly


dilated


slit)


illustrated


Gruber


(1967)


for juvenile


lemon


sharks


after


0-3 min


of dark adaptation


following


intense


light


adaptation.


In the


course


of dark


adaptation


lemon


sharks,


there


is less


than


a one log


unit


change


in sensitivity


within


the first


three


minutes,


compared


with


an ultimate


adaptive


change


of six log


units


when


fully


dark-adapted


(Gruber,


1975).


Further-


more,


retinal


illuminance


in the mapping


experiments


given


a pupillary


area


of 18


2. a medial


("posterior")


nodal


distance


of 9


.398


mm in the


juvenile


lemon


shark


eye (Hueter,


1980),


and luminance


of 136.4


cd/mm


was estimated


to be well


over


7000


trolands


(Brown


et al., 1987)


which


is beyond


the region


of rod saturation


for human


vision


(Wyszecki


Stiles,


1967).


Thus


seems


unlikely


that


the shark


became


scotopic


under


the conditions


of the mapping


experiment.


Visual


stimuli


eliciting


the best


tectal


responses


for mappin


pur-


poses


were


circular


black


discs


with


diameters


subtending


or 10


visual


angle when


positioned


on the hemispheric


surface.


The discs


were


mounted


on clear


glass


rods,


so that


they


could


be manipulated


freely


over


the hemisphere


from well


outside


a unit's


receptive


field.


Also


occasionally


used


in delimiting


the ed


receptive


fields


were


black


card


mounted


on a glass


rod or a black


bar.


Under


routine


condi-


tions,


however


the discs


alone were


used


to locate


and define


recep-


tive


field


positions


in the mapping


experiments.


Electrodes


and Instrumentation









from


Insl-x,


were


insect


based


opened


pins


upon


(BioQuip)


the method


immersion


them


electrolytically


of Green


sharpened


(1958).


an instant


and insulated


necessary,


in ethanol,


until


with


the tips


a resis-


tance


of approximately


7 M2 was


obtained.


The sharpened


insulated


pins


were mounted


in 23-gauge


syringe


needles


(with


connectors


removed)


gold-over-nickel/


copper miniature


connectors


(Amphenol)


were


crimped


to the needles.


Electrodes


were


tested


recording


from


teleost


tectum


before


conducting


the shark


experiments.


use of


metal-filled


glass


micropipettes


as mapping


electrodes was


also


attempted,


but these


were


ineff


ective


in consistently


penetrating


the relatively


tough meningeal


layer


overlying


the shark


tectum,


and thus


they were


unsuitable


lengthy mapping


experiments.


stainless


steel microelectrodes


were mounted


a hydraulic


microdrive


with


digital


display


(Haer)


which


provided


remotely


con-


trolled


stepping


of electrode


depth


z-position)


at steps


of 10


The microdrive


head was


mounted


in a manual


x-y micromanipulator


(Prior),


calibrated


in 0.1


mm steps,


which was


used


to set the position


of the electrode


calibrated


across


movements


the tectal


of the electrode


surface.


This


in three


arrangement


planes


(x-y-z).


allowed

The


orientation


of the electrode


and shark were


aligned


with


respect


to each


other


so that


the electrode was


perpendicular


to the dorsal


surface


tectum,


and the


movement


the electrode was


parallel


to the


sagittal


transverse


planes


through


the brain.


After proper


align-


ment,


the electrode


mount


was firmly


anchored


in place


to the laboratory


1 __


I









monitored


through


this


microscope


as a periodic


check


on the animal's


condition


during


the experiment.


After


ocular,


tectal,


and electrode


orientations


were


secured


proper


ference


(rostral,


with


alignment,


used


purposes,


caudal


the electrode


recorded.


Second,


the electrode


ways.


medial


tip,


First,


, lateral)


and their


selected


to plot


the four


the right


corresponding


a distinctive,


tectal


position


dimensional


tectum were


re-


limits


located


coordinates


easily


identifiable


were


melano-


phore


from


pattern


such cells


distributed


over


the tectal


surface,


and plotted


its position


with


the electrode,


Jacobson


(196


did for


reference


purposes


in the frog


tectum.


These


five


reference


points


could


be reconfirmed


throughout


course


of the


experiment


as a check


on head


receptive


movement.


field


Furthermore,


was


plotted


the melanophore


on the hemisphere


site's


corresponding


as a reference


field,


periodically


returned


the electrode


to that


site


to confirm


that


corresponding


field


position


not changed.


Conventional


amplification,


audio-monitoring,


and visual


display


methods were


used


to record


unit


activity


Single-ended


input


was used


from


the electrode


to an AC


preamplifier


(Grass


P15)


mounted


near


preparation


and grounded


to the


seawater


bath,


the metal


retractor


the cranium,


the micromanipulator,


and the binocular microscope.


preamplifier


gain was


and the low cutoff


set at 1000X


filter


the high


at 100 Hz.


cutoff


output


filter


signal


at 10 kHz,


from


pream-


plifier was


passed


through


a 60-Hz


T notch


filter


and then


split


lniii n


mni-i tfn


(Cr'nsc


AMNL


n nrnnsnra


osri 11 nnnen


(Tlekt ronix


502A)-









procedure.


permanent


records,


photographs


were


taken


from


storage


oscilloscope


screen


with


an oscilloscope


camera


(Tektronix


C-5A),


using


output


from a


silicon


photocell


to track movement


of the pho-


tic stimulus


through


a unit


s receptive


field.


Mapping


Procedure


based


my mapping


technique


on that


developed


Schwassmann


Kruger


(1965a,b)


and Schwassmann


(1968


1975)


for teleost


fishes.


After


setting


the desired


position


of the electrode,


was lowered


to the


tectal

tor.


surface


When


by manually


the tip made


cranking


contact,


down


switched


stage


of the micromanipula-


to the microdrive


unit


slowly


penetrate


tectum.


As the electrode


was advanced


through


tectum perpendicular


to the dorsal


surface,


stimulated


the left


with


a broad


photic


stimulus,


usually


a moving


flashlight


beam,


until


the level


of maximum


unit


activity


in the


tectum was


obtained.


The spatial

was affected--the


area

unit


on the hemispheric


'


receptive


surface


field--was


in which


delimited


unit


with


activity


the dis-


create


photic


stimuli


described


previously.


boundaries


of this


field were


determined


by moving


the stimulus


or 5


black


circular


disc


against


a white


background)


in small


steps


over


the hemispheric


surface,


and marking


outer


edges


the responsive


area with


colored


pen.


This


area,


outside


of which


no response


of the unit was


observed,


constituted


the receptive


field


for the electrode


position at


specific


unit


x-y-z


coordinates,


number matching


that


and these


coordinates were


of the corresponding


delimited


recorded with a


field on t


he









In this


way,


the dorsal


tectal


surface


was


mapped


in 0.5


mm steps.


a complete


mapping


experiment,


a grid


of equally


spaced


electrode


sites,


each


mm away


from


next


site


along


a rostral-caudal


medial-lateral


transect,


is obtained.


mm tectal


surface


potentially


yield


to 150 such


recording


sites,


but due to time


con-


straints


this


was impractical.


The maximum number


of surface


sites


that


successfully


recorded


from


in one experiment was


and that


experi-


ment


ran for


over


24 hr.


Most


of the successful


mapping


experiments


for 10-14


As with


ual field


and this


was


teleosts


found


region was


(Schwassmann,


to project


mapped


1975)


to the lateral


penetrating


a portion


wall


of the ventral


of the shark


the electrode


beyond


vis-


tectum,


overlyin


surface


sites


to these


deeper


units.


The medio-lateral


and rostro-caudal


coordinates


for these


deep


sites


were


same


as for the overlying


sites,


and their


depths


below


the tectal


surface


were


recorded.


Special


characteristics


of receptive


fields


such


as directional


selectivity


occasionally


recorded


were


and unit


noted


multi-unit


not rigidly


experiments


classification


during mappi

responses,


standardized,


or to other


studies


into


Since


the steel


and the photic


the comparability

is diminished.


stimuli


or ON-OFF


units


electrodes


were


primarily


and adaptive


of these

As long a


data


state


between


s spatial


orien-


station


of the electrode


cal mapping


criterion


and preparation


of the


is preserved,


center-to-center


however,


distribution


the criti-


of receptive


fields


can be accurately


assessed.


n f trcho ov-rTnont' r


C^tln QO n


con-


can


ran


thlh noo rlo


T~ii r- r n r


~nrrcITC


I I


I


rI









Since


curare


is metabolized,


this was


sometimes


necessary


about


every


standing


anesthetic


level


of 1:20,000


tricaine/benzocaine was


sufficient


throughout


even


the longest


experiment.


These


drugs


have


additional


bleeding,


determination


benefits


reducing


of receptive


of decreasing


spontaneous


field


blood


pressure,


activity


boundaries


and thus


of visual


more


units,


precise.


amount


making


On the other


hand,


tricaine


and benzocaine


have


been


shown


to adversely


affect


reti-


function


inhibiting


dark adaptation


in frogs


(Rapp


and Basinger,


1982;


Hoffman


even


and Basinger,


juvenile


lemon


1977),


sharks


newts


(Hamasaki


(LaTouche


and Bridges


and Kimeldorf,


1965),


1978),


probably


blocking


the regeneration


of rhodopsin


(Rapp


and Basinger,


1982).


There


no evidence,


however,


that


these


drugs


can affect


the topographic


lationships


in the retinotectal


proj section


pattern.


At the conclusion


of the mapping


experiment,


the receptive


fields


plotted


on the hemisphere were


transcribed


a data


sheet,


on which


printed


and the


a two-dimensional


coordinates


representation


the hemispheric


of corresponding electrode


surface,


positions were


plotted


on graph


paper.


Histology


Fixation


Experimental


animals


to be sacrificed


for histology were maintained


under


anesthesia and


dark-adapted


for approximately


hr prior


to enucle-


re-


was









A small


quantity


of 10% formalin-saline


was injected


into


the vitreous


of each


eye through


the optic


disc


(Hebel,


1976).


The saline


used


this


fixative was


formulated


an elasmobranch


.H. Evans


(after


balanced


Forster


saline


et al.,


("Ringer


1972),


solution


and is shown


Table


This


fixative


yielded


excellent


results


exceeding


those


typi-


cally


obtained


with


formalin


fixation.


eyes


were


immersed


in 10% formalin-saline,


where


they


remained


for at least


several


weeks


in fixative


before


dissection.


Natural


land-


marks


in the


eyes


pupillaryy


shape,


optic


disc


location)


allowed


identification


of 1


vs.


right


correct


orientation


of each


during


later


dissection


the retina.


Retinal


Wholemounts


Techniques


for preparing


retinal


wholemounts


(flatmounts)


were


adapted


from


guidelines


described


Stone


(1981)


After


fixation,


was opened


at the corneo-scleral


margin,


and the retina


was


dissected


free


from


the retinal


epithelium


in elasmobranch


Ringer's.


preserve


spatial


orientation,


small


knicks


were


made


in the dorsal-most


ventral-most


dorsal


retina


and ventral


prior


pupillary


to removing


corners


anterior


as guides.


segment,


isolated


using


retina


then


floated


onto


a large,


gelatinized,


glass


covers lip


x 65


mm),


radially


cut a minimum number


places


around


the border


achieve


flattening


of the retina


onto


the coverslip.


With


a fine


camel's


hair


down


brush


onto


and small


the coverslip


of filter


photoreceptorr


paper,


layer


gently


down)


brushed


and removed


the retina


excess


was


was









Table


Elasmobranch


(after


Forster


balanced


et al.,


saline
1972).


(Ringer's)


solution


of D.H.


Evans


NaCI
KC1


CaC12
MgC12


16.35


*2H20
*6H20


Na2SO4
NaH2PO4 2H20
NaHCO3


Urea


21.00


Trimethylamine


oxide


(TMAO)


Glucose


Polyvinylpyrrolidone


(PVP)


30.00









does


not induce


the artifacts


tissue


shrinka


that


are produced


with


xylene-clearing


and alcohol-dehydration


methods


(Grace


and Llinas


, 1985).


Curcio


et al.


(1987)


have


recently


described


a technique


for using


DMSO-


clearing


of retinal


wholemounts


to analyze


photoreceptor


gang


lion


cell


topography.


utilized


their


technique


with


some


modifications


follows.


Several


drops


of DMSO were


placed


on the flattened


retina,


a second


large,


but nongelatinized


coverslip


was positioned


over


tissue.


This "sandwich"''


was


then


placed


a dish


containing


DMSO


with


a small


weight


(about


100 gm)


sitting


on top


of the dish


lid and lightly


pressing


down


on the sandwich


to maintain


flattening.


The retina


allowed


to clear


in DMSO


for about


three


days.


After


clearing,


the overlying


coverslip


was carefully


separated


from


retina,


excess


DMSO was


removed,


and the retina was


saturated


with


glycerol.


sandwich


were


new


large


sealed with


coverslip


lacquer,


was


applied,


resulting


and the edges


a cleared,


of the


nondehydrated,


unstained


wholemount with


relatively


shelf


life


(Curcio


et al.,


1987).


Shrinkage


of the tissue


using


this


technique


minimal


com-


pared with


the approximately


10% linear


shrinkage


usually


associated


with


lene-alcohol


techniques.


In fact,


Curcio


et al.


(1987)


reported


a small


net expansion


of 3.2%


, primarily


a water


rinse


step


which


omitted.


Recently


Dawson


et al.


(1987)


have


utilized


same


flatmounting


technique


as mine


to count


ganglion


cells


in human


retina,


and they


found


negligible


shrinkage


of the tissue,


based


upon


the size


retention


of erythrocytes


in the finished


flatmount.


' 1% T JT n 1n I rnn iT, + n A


ud f. n rscr an


..... o _


--__* 1


r-


n


,, Pa? r b rt


R; ~,n T t-


was


mI I -


n j -ir- yi a~ ,-- T. 1









eyepiece


contained


a linear


scale


calibrated


with a


100 um


calibration


slide.


The other


10X eyepiece


contained a


cross-hair


reticle


dividing


the field


into


four


quadrants.


The microscope


stage


was


equipped


with


scales


for determining viewing


location


on the tissue


to the


nearest


0.1 mm.


An outline map


was made


of each


retina


on graph


paper,


using


microscope


and calibrated


scales


to trace


the coordinates


of the flat-


mount


features.


This


map was


then


used


to specify


sampling


sites


which


photoreceptor


ganglion


cell


counts


were


subsequently made,


and these


data were


recorded


at their


corresponding


locations


on the


map.


Orientation


of the retina


was


preserved


using


the dorsal


and ventral


cuts


and optic


disc


as landmarks.


Cell


counts


were


converted


cells/mm


and isodensity


contours were


drawn


following


the guidelines


of Stone


(1981).

















RESULTS







Retinotectal Mapping


Responsive


Layer


Receptive


Fields


Retinotectal


projection


maps,


either


complete


or partial,


were


corded


with


20 animals.


Initially,


success


in getting


a shark


through


surgery


and obtaining


a usable


was low


, less


than


50%.


But after


gaining

control


considerable

of bleeding)


experience


depth


with


the surgical


of anesthesia,


choice


preparation

of electrode


(especially

. and the


electrophysiological


technique,


the experimental


procedure


became


routine.


Recording


depth


maximum


unitary


response


to photic


stimuli


subject


to variables


which


include


size


of the animal,


specific


location


on the te

electrode


ctum,


and the degree


is advanced


of dimpling


(Schwassmann,


1968).


the tectal


The last


surface


factor


as the


is influenced


the shape


of the electrode


and the toughness


of the meninges


over-


lying


the tectal


surface.


attempted


to circumvent


the problems


of dim-


pling


resettin


the electrode


to the tectal


surface


after


each


initial


penetration,


With


and then


this


beginning


adjustment,


depth


average


measurements


depth


from


of maximum


that


point.


response


for all


A-1- -- r -


re-


/" r I, -i-t -^ -


C


L


I_


*


__1_


LI-~- --L-- r


--a


/.


Y


n









Typical


discharges


of these


"tectal


units"


In response


to photic


stimuli


are shown


in Fig.


Signal-to-noise


ratio was


acceptable


mapping


purposes,


as typical


recorded


responses


were


approximately


50 IV


bursts


over


of 70-80


a background


were


recorded;


noise


of about


rarely


these


10 iV.


Occasionally,


discharges


reached


dischar


100-120 lV


some


isolated


tectal


units.


With


the steel


microelectrodes,


great


majority


of recordings were


not from


isolated


single


units


rather were


multi-unit


bursts


, pooling


responses


or more


units with


overlapping


receptive


fields


(RF's).


This


limits


the effi-


cacy


of determining


specific


RF properties


beyond


the mapping


criterion


center-to-center


distribution.


Furthermore


such


factors


as adaptive


state


of the animal,


background


luminance,


and depth


anesthesia


influence


the properties


of visual


receptive


fields.


With


this


proviso


the following RF


characteristics


were


noted.


OFF units


predominated


but examples


of ON-OFF


and ON units were


also


found.


Strong


directional


selectivity


of stimulus


movement


through


unit


s RF was


occasionally


observed


and noted,


although


time


constraints


of the mapping


experiments


would


not permit


a systematic


unit-by-unit


evaluation.


However,


in those


few units


analyzed,


no particular


pattern


directional


selectivity with


RF location


in the visual


field


found.


On the other


hand


location


in the visual


field


did affect


the size


and shape


of tectal


unit


Far peripheral


RF's


beyond


about


the dorsal,


caudal,


and ventral


visual


fields were


larger


or occa-


a


- .4


-. 2 -nil -- -.. 1


was


1 1


Lrrl


1


q -. A A


~


1


LLn- rr-lurll~

























A A A A


A A A A A


20 IV


500
msec


d-


AA AA A A A A A A


V AY AV AT AT


Fig.


xtracellular


tectum


recordings


response


from


to photic


tectal


units


stimuli


in lemon


each


case


shark
, the


trace


stimulus


shows
change


unit


in the


discharges,


unit's


the bottom


receptive


field


trace


shows


as monitored


a photocell.


, b,


units


are centrally


located


black


units


disc


firing


through


their


response t
receptive


o the
field


passage


(arrows)


.*


-- T









(visual


horizon),


however,


unit


RF's


took


on a pronounced


elliptical


shape,


with


vertical


minor


axes


subtending


about


or less,


and hori-


zontal


major


axes


subtending


about


7-10


Retinotectal


Topography


General


projection


pattern.


many


as 97 different


tectal


units


were


plotted


in the


most


complete


mapping


experiment


on a


single


shark.


Even


in partial mapping


experiments,


the general


topographic


orientation


of the visual


field


map as superimposed


on the contralateral


tectum was


clear:


the dorsal


visual


field


projects


to the medio-dorsal


tectal


surface,


the rostral


(anterior,


nasal)


VF projects


to the rostro-dorsal


tectal


surface,


the caudal


(posterior


, temporal)


VF projects


to the


caudo-dorsal


tectal


surface,


and the ventral


VF projects


to the lateral


tectal


surface.


visualize


this,


it is useful


to think


of the hemi-


spheric


bubble


that


simulates


the left


visual


field,


turn


it inside


out,


and overlay


it with


that


orientation


on the right


optic


tectum.)


With


the inversion


of the visual


field


on the retina,


this


means


that


dorsal

jects


retina


projects


to the medial


to the lateral


tectum,


the rostra


tectum, t

1 (nasal)


:he ventral

retina pr


retina


objects


pro-

to the


caudal


tectum,


the caudal


(temporal)


retina


projects


to the rostral


tectum.


Topographic


maps.


Seven


selected


examples


of actual


data


map s


from


the 20 experiments


are shown


in Figs.


4-10.


Each


figure


shows


the left


visual


field


as a plane


projection


of the hemispheric


surface.


shark' s


right


tectum is


shown


in dorsal


view,


with medio-lateral


(M-L)











EC a)

oo
I00 0
0 ~ U- J Jlt
u d ad TO^

O 03- *v- 0 )

03 CO c 4
C' *HC 0
TO 1 4Oia O -


03O* *-J Te
*r- -4U2r- c

r-4 n 0 *rl 4- '0 S
i+- Ci ^-1 M- C Cl
jC 4 2 0C

)rn V) 1-1 c:
-t CO tF 4- O--
*H o *H u


(dNc aE-a


-- 0 V .0 1 W C
to a c






3 C5ri- TO4W^0)
-rl U* -I ( 0 -i
> t OU' TO*







COW 0-


,r040 4- C 04
04-1 J- O-O C
ul.r Cu 0 *
QJ M-i N (
r- 0Ps 't3 T- "0J
Ii-i ^ .-
H ^: l- O-
,Q, WO QI *
Curi v 4i UC
GC (fl 4-

0c *H M -i ) *(
*lf T H TO* O
v 1r 3 ir- a r x

^ E*4" *H cu
2J ~3 >d TO 6 r
Su a< u
0) Co o r= Ti

l *1Q 3
/'-N U CJ-
uC i: *^ C
vau 0>
O *H USi .r- (U


,0 3 40



0 3


44 0


U 0 T U2. 0

0 4~

4 C) CO


O 0 H 0



0 O. .T-' 4-
*irl 0 EH QJ 0


0 -t ,


-RQ ) 0 403
SQCNA
*ifl GO sU Q
C C C

TO 4 W 4
40 *QO U

EH iOJ34


C.
TO
S


> OO TO



0o w U '0
0 O T CO T Q
a 0 C 0 S-i
C *ii0 i-i CU LU
.,-4 4-Jr CsW& 4 ^ C
TOl U ^C ct 0)
CU4Q C ii1tC
WOOrH OWM
4-000

AJ i-0 (U *l C E
C* g O E
vJ (U L 1 h (
1U 0' 1rU -

a i-t CJ TO p J-J
0) Q4oJ 4 Or'
Qs4-JW( 34-1 C .
0 Ot0) U OCO<
k ,-i >CO *H(

S Q*- xCL A-
40r403r



AJCr cs 1 a !
W(cU )U
EJ *r, i : ,D




EtMW ^ *
uC-r
T O &- c 1X
.a (Usi U Q
CO X: I QJ &. rr-

0C0 U 0>0
TO Ga u0 c

,= C QJ TO u- -
i-i n o c~
Fi 3r &0D 3- r 6)
0 Xl C CO Cas '
0 *r'
U) WrOrTO TO 3
N 4J 44 3
*r- -U) O i *
*I- Ir CO VU TO *
CO03 r 3
(Ur QJ >: T-l 4
3 C 3- ?-<
*1034 (U(- J s} 4C

U^lr Cv >J 7M
4-J4- 'H C'-U2 CO
.0 0 0T
,OI O O*H









SM 0o nt U
4 aTOO
CMu ^ c Q


o
Oi


CC t*TLOC 4444
S0 *H0 U 0)
040 H>C 0O
-O C C i*- 0 0
O 4W TO C QJ- -O



E-U) SCci 4. r- Cn i
*C <(T000
*u. 0 U ,0
** O 4 QC r- -I- r4
*H C 4TO C 34 40 0
00* *U *4
U0 *4 1 4->UCT




U 0 3( U ci ( O
3 C U 3 O > 0
u ( ( 0 0 0 QJ r- i-





O


/ ^\///#/W/
















'C

-o
0)



4-a
Irl



ytl


rlt



ow
3


r-l
>





2-4
W3
d

91

C)

4-a
I C


0 )









a~
4J e
C















S

4 J3
0 0
















SW
0-I
(U
C U


0. 0



4J
U i




















( Cn
*ri C






0' -
co o


cu



4- QI

*!1 *l l

4-) en
QJ 0


-w Or

act




g u





O


W////////////A

W4//////////A~Q

W//////////A
Y%//////























u n Oc
4o eoD 0


) a) 0 ( 0
.n *o *
.,0 0 *0


4-t3 (U
00..
QJ cd o en ^
mo E

MO 0 *
40 H0
u *



am
rl IM
'rC CL

MCO QJ \
00 I


0 0- 3
U) 4-1 E -
.J) '^C' CM\C
u N Cr


n3~ cu E:
ca cu (U c
4-J $-4-J 4J4J

a3 Q) 0)
-o C3 c 3
0 w w

o i-< o a)

4-|i (t CO
Q) 'O r- IC1 G-

F0 > & ^
u W r-4
U 3


I W4- cu e
g oo -I 4.2 N
c in -
e-d O- 0rH- 02
~icn ^ cZ
w-4.t OnCL
4jE- $*4'V4-^










\^^^^~/////^{











C
c v
0

T OJ
*1-


&. ^00
rd ac
E o
GO
E0
Cfl.


'0
0) -o
U*
(tt1n- u
CO0 0c
*1 1C.
ar ,


0)

T? en t3
SO 0- C
Sa to
ka e-4
4- co

-41 o

0 *


0 N r-4
003


C- C
a bs
?< c
QJ *H- ..C

X C3
0c3
0 00
4-C M/X



e co 0
4-t vlC



t3 C L4.<
oQ C0





68

O








.C~
0-


N
a-

LL

















0W \
u-~ \Y I r I

















































#t

ow0
O q
N
4J erl
oC w





70
0




CO
a)
a.


^~ I I /
U,

N

LL


-*




\ ~Y I /
u.~ "-4 --
U= ,




U-4
U.








0 _m

















S

-i

a
C
o



rl)
'0
rJ



Crl

PI

cr y





C 'l


4-i


a) a
00)


U) Cl)



lCM
I
^O













C)J
r- Cl
o -4


C-'









a a)


Ik
0)0
-'-44P













a:ar
Qj(U






















04-4
7 CD



C/ CO























a,
CJ




CO n



4-o
CO Cw
rCO
o
a)'
Nd u-
^ (u



Q) O
(d m
^ *r





72



Q.)



c t-j # 9
,C


N-
a.









\m
QC CM














t.4
U- x i.
c o












INJ















c
:3
0
o





cn
0
e3

(0








-z

(4J

Co
02
4-a
3




U
CU
411


+J





0)


Ce






C)




QI

nao
4-1 V]





dU2
*-z
t-i en

03 o








0)
en c



ow


ca &












oSco
CO
:3
b, a
cn


Co>
N











E44
a,





E 1




E -a

e
Fz u


ctD


4-t CU
ca ct





74
O





cu


C '









With


few exceptions


where


indicated


on the figures,


neighboring


electrode


sites


along


a single


M-L or


transect


were


spaced


an equi-


distant


mm from


each


other.


This


means


that


absolute


shifts


in RF


position,


measured


in degrees


of visual


angle,


are spatially


comparable


between


electrode


sites


along


a given


transect,


and between


transects.


Thus,


nonlinearities


in the shark


retinotectal


projection


would


expressed


ters


as a compression


corresponding


or expansion


equidistant


of the


recording


spacing


sites.


between


A totally


cen-


linear


projection,


then,


would


reveal


a completely


uniform


distribution


of RF


positions


any nonuniformity


in this


pattern


would


be indicative


a nonlinear


projection.


In Fi


4-6,


only


the RF


center


positions


as represented


cir-


cular


fields


with


numbers


corresponding


to the


accompanying


tectal


cording


sites,


are mapped.


Fig.


shows


the results


most


complete


single


mapping


experiment.


RF positions


as well


as sizes


and shapes,


determined


under


experimental


conditions


were


mapped


in the


experiments


depicted


in Figs.


7-10.


In each


case,


the optic


disc


(O.D.)


was located


and mapped


shown


ophthalmoscope,


in the dorso-caudal


and the disc


quadrant


s corresponding


of the visual


location


field.


The first

nonorthogonal


characteristic


nature


that


of the RF position


apparent

ns as the


in the data maps


tectum


is the


is traversed.


Proceeding


from medial


to lateral


tectum


there


a movement


corres-


ponding


RF locations


that


first


descends


a slightly


rostral


direction,


but then


ab rup tly


twists


forward


as the central


visual field


n-- 1 1 n


/fr -


re-


- .C 0


"7\


4- 1- -tfj- jr


,,,,,1,1


Y~nnnn: *


I. C-









The forward


descent


of RF positions


with


medio-lateral


movement


across


tectum


indicates


that


the visual


field


represented


the tectal


ittal


surface


transverse


is rotated


planes


about


of the


clockwise


tectum.


from


In other words


the precise


, my M-L


sag-


elec-


trode


transects


were


oriented


about


off the representative


field


lines


overlying


the tectal


surface.


These lines


apparently


run somewhat


caudally when


itself has


proceeding


no functional


from


the medial


importance


to the lateral


terms


of spatial


tectum.


visual


This


organiza-


tion,


for the


precise


alignment


the tectal map


not indicative


nonlinear


specialization,


mere


a developmental


result.


But the


nonorthogonal


twisting


RF positions


through


the central


visual


field,


across


the rostro-caudal


horizontal


meridian,


hints


some


sort


non-


linearity


in this


region.


Returning


to the dorsal


visual


field


(Fig.


the RF posi-


tions


are spaced


relatively


apart,


as much


as 20


but typically


about


or less


mm shift


uniform with


on the


equidistant


tectum,


sampling


and their


of the medial


distribution


tectal


more


surface.


Thus,


the dorsal


visual


field


is projected


linearly


on the lemon


shark


tectum.


The ventral


visual


field


(Figs.


4-10)


is similar


to the dorsal


field


in that


the projection


also


appears


to be relatively


linear with


widely


spaced


receptive


fields.


This


is a much more


difficult


region


map,


however,


due to the location


of corresponding


tectal


units


(numbered


with


subscripts


a-d in Figs.


4-10)


on the rostro-lateral


wall


tec-


turn rather


than


on the dorsal.


tprtal


surf are


Tn reach


theseP


sites.


: a


L









This


difference


in sampling


technique


most


likely


introduces


small


spatial


distortion


in the ventral


map,


extent


of which


is dif-


ficult


to correct


for exactly,


but several


points


should


be considered.


First


, although


some


workers


have


attempted


to correct


for tectal


curva-


ture


Siminoff


et al.,


1966),


this


correction


has been


relatively


minor


and has


not significantly


changed


the overall


projection


pattern,


leading


subsequent


workers


(e.g.


Schwassmann,


1968)


to reject


the need


for correction.


Second,


the shark


tectum


does


not turn


under


curve


medially


under


the optic


ventricles,


as in teleosts


(Schwassmann,


1975)


rather


ends


on the lateral


wall.


This


minimizes


the positional


sampling


error


of the deep


rostro-lateral


sites


in shark


tectum


compared


with


teleost


tectum,


the ventral


and it also


visual


fields


explains


of the lemon


differences


shark


vs.


in the orde

teleosts.


of RF's


For example,


in Fig.


the order


of RF's descending


through


the ventral


field


cor-


responds


first


to tectal


unit


#6 (the


next-to-last


surface


point)


then


#7 (last


surface


point)


then


#6a (in


this


case,


1.04


mm below


mm below


#7),


(0.90 mm


below


#6a),


and finally


(1.00


below


#7a).


The comparable


order


a teleost


would


be 6-7-7a-6a


(with

tectum


no deeper "b"

(compare with


points),


because


the serranid


of the infolding

in Schwassmann,


of the the teleost

1968).


Finally,


sites


in reconstructing


as represented


the


on the shark's


actual spacing

rostro-lateral


deep


tectal


recording

surface, I


found


that


geometric


distortion was minimized


because


the deep


points


A- I -- 1---1 -1 .--? -- C .- -~
S
- -- ----- -~


I


,..,r,,,


_ -_ I __ i


,,1,~,


1


___L_









In consideration


of these


various


points,


chose


to incorporate


the ventral


RF positions


into


the composite


retinotectal


as is,


with-


out correction.


No retinotectal


magnification


or other nonlinearity


therefore


detected


However,


there


in the lemon

is obvious


shark's


lower ventral


magnification


present


field.

in the horizontal


expanse


of the central


visual


field,


which


projects


to the central


dor-


sal roof


of the tectum where


no curvature


distortion


is involved.


Figs.


4-7 and 10,


a clear


compression


of RF positions


can be


seen


in the


horizontal


band


running


approximately


above


and 150


below


the hori-


zontal


meridian.


Within


this


band,


where


the RF


procession


twists


accommodate


greater


input


of tectal


units?)


spacing


succes-


sive


centers


is about


which


is about


one-third


of the visual


angle


subtended


the spacing


between


the most


peripheral


units.


This


band


appears


to extend


unbroken


from nearly


rostrally


(Fig.


caudally


(Fig.


, although


specification


of the


extreme


rostral


is compromised


the difficulties


of reaching


the corresponding


proj ec-


tion


sites


on the rostral


tectal


wall.


only


side


are there


the band,


differences


but the sizes


in RF spacing


and shapes


between


receptive


inside


fields


out-


appear


be different


as well


(Figs.


7-10)


, as previously mentioned.


Peripheral


units


have


large,


somewhat


irregular


receptive


fields,


whereas


units


within

50 or


the band


less


have notably


vertically


and 7-1


smaller receptive

0 horizontally,


fields subtending about

resulting in a character-


istic


elliptical


shape


oriented


parallel


to the band


itself.


These


L -- -


--,,-- -


- -. ..- 2- L .-- -


--,, -


,_l---l_-L I----


,-, Lu,









circular


center


and elliptical


border


(weaker


"hash")


vs. outside


of the


ellipse


response).


Composite


maps.


Figs.


11-14


represent


an attempt


to pool


suits


of all 20 mapping


experiments


into


a series


of composite


maps.


These


maps


incorporate


only


positional


data


of the RF


centers


plotted


the 20 experiments,


as these


RF's


would


project


to the


theoretical


"aver-


shark


face


tectum measuring


and 32 subsurface


mm x 5.5


recording


mm (R-C


sites


x M-L)


11) .


, yielding


composite


sur-


maps


were


produced


piecing


together


the standardized


results


of the


most


complete


mapping


experiments


in order


summarize


the general


projec-


tion


pattern.


12 shows


the left


visual


field


with


the complete


representa-


tion


centers


corresponding


to the electrode


locations


plotted


Fig.


border


tectal


The single


between


surface


area


and the


curving


line


of the visu


area


in Fig.

al field


projecting


12 marks t

projectin


he approximate


to the dorsal


to the rostro-lateral


wall.


stippled


zone


encloses


the horizontal


band


compression


of the


recep-


tive


fields.


connecting


the RF


centers


plotted


in Fig.


12--in


other words


, by


reconstructing


the theoretical


and R-C electrode


transects


as in an


ideal


mappin


experiment--the


shown


in Fig.


13 is obtained.


The 100


visual


field


lines


have


been


removed


from


this


to avoid


confusion.


In this

through


figure,


the nonlinearity


horizontal


band


in the transition


can be clearly


of the RF distribution


seen.


Finally.


transferring


the 100


visual


field


lines


of Fie.


re-


-A-J-


I V


UV^-













Latera


* Tectal surface point
O Deep rostro-lateral point


Medial


1 mm


Fig.


Dorsal
cover


view


of right
complete


optic


tectum


juvenile


0.5 mm-spaced


lemon


electrode


shark
sites.


The unit
Fig. 12.
lateral


numbers


The
tectal


correspond


curved
wall.


line
The


to the RF


represents
X (unit 67


centers


plotted


the underlying
is the site r


rostro-
eceiving


input


from


center


of the visual


field


(geometric


center


of the retina)










Dorsal


9l



2a2.
2 29 2




10 S
@2 a
130






Sa ,2 1.t ~
SII


@262


*7g 9 -- ^ f' 2 --
PC-

1sa ?2 gT\
102 29
$112 135
992 136c~
1220
10



113
137

102a123 L131
1212


1013a 123a
113a


Ventral


Fig.


Composite map


of the locations


of RF


centers


in the left


visual


spending


field


of the lemon


electrode


sites


shark


plotted


proje


acting


to the


in Fi


corre-
approx-


imate


border


between


surface and deep
scripts a or b)
horizontal band


units


projecting


rostro-lateral


is shown


units


the single,


of RF compression


to the
(number


curvin


is indicated


dorsal


with
line.


within


tectal


sub-
The
150


A4 nc 11 i nr1 nnrA


TnQt- ri -T ^a


TT/^S t 4- "t


l


l


I


)27

)238




49IJ 5 I ;:!
wl :3 f
: ss-:-:::- ulq ,60
28 71m


25a
28a


g~36 36 47 58a



47b59b /7

81b

91b


/I ^-\Vl- I


In -- I


<-/- +-T-1










Dorsal


Ventral


Fig.


Composite


centers


visual


plotted


continuous


sects


of Fi


field map


lines


in Fig.


sampling
Except


These


along


connecting


lines


the M-L and


for the horizontal


neighboring


represent
R-C tectal
and verti


ideal-
tran-


meridians,
confusion.


the 100


field lines


The concentration


have


been


of tectal


removed


to avoid


representation


into


the horizontal


band


within


15D


15V


is clearly


seen


r


r


\ t/:_


j~~


: ?::(~~~ ~I
%: :: : '~:: ::~:~: .


I


.~:~ :~:::::-
.. ::
5 i;::-::ji::,:~ ;
-:^~E~::'::;6:8 J~.:~:i
:::~:::,:,: ::~ iii
:~ ,:,,
~~~::~j4 t!i:"':
:2~j:i~:-ji
F: ::k~













Lateral


0 D-





150D


OR-C


15V


00D-V


OR-C


Medial


Fig.


Retinotectal


projection


the left


visual


field


onto


surface
shark.


of the right


Horizontal


optic


tectum


and vertical


in the juvenile


meridians


lemon


are labeled


0D-V


by an X,
rostral


orsa


and 0R-C,


represent


and ventral


tectal


respectively.
ts the visual


visual


surface


Their
field


field


onto


intersection
center. The


shown


wrapping


the rostro-lateral


, marked
extreme


under
tectal


wall.


Retinotectal magnification


the visual


streak


between
surface


15D


and 15V


is devoted


is such


to this


that


visual


52% of the


area


tota


representing


tectal


only


26% of the total


visual


field.









the right


optic


tectum is


obtained


(Fig.


14).


the retinotectal


mag-


unification


of the visual


field


contained


within


the 15D


and 150V


field


lines,


hereafter


referred


as the "visual


streak,


is clearly


evident.


Calculations


show


that


this


visual


streak,


which


comprises


only


26% of


the entire


hemispherical


visual


field,


is represented


available


tectal


surface,


including


the rostro-lateral


wall.


mining


74% of the visual


field,


hereafter


referred


as the "peripheral


field,


is represented


only


48% of the tectal


surface.


This


means


that


ratio


of proportional


representation


on the tectum of


the visual


streak


52/26


= 2.00)


vs.


the peripheral


field


(48/74


= 0.65)


00/0


= 3.08.


Thus,


averaged


across


the entire


visual


field


about


three


times


more


tectal


surface


area


receives


input


from


units within


the visual


streak


than


from equivalent


areas


of the peripheral


visual


field.


Retinotectal magnification


factor.


This


concept


of proportional


representation


typically


expressed


terms


of the retinotectal


magni-


fiction


factor


or RMF)


(not


calculated


to be confused with


as the number


the retinal


of micrometers


magnification

on the tectal


fac-

sur-


face


per degree


of visual


field


coverage


(Jacohson ._196


* For most


peripheral


dorsal


and ventral


projection


areas


units


and 15 in


Fig.


this


number


averages


about


500 pm/150


= 33 Um/o


For the visual


streak,


this


number


is 500 nm/50


= 100 pm/


Thus


the ratio


of retino-


tectal


magnification


factors


is in agreement with


the surface area


com-


putations:


three


times more


tectum is


devoted


to vision


in the visual


re-









Retinal


Topography


Photoreceptor


and Ganglion


Cell


Identification


Results


of the DMSO-cleared


retinal


wholemounts


combined


with


NDIC


microscopy were


generally


excellent


for the photoreceptors


and within


acceptable


limits


for the ganglion


cells.


These


results


far surpassed


my previous


attempts


to visualize


photoreceptors


and ganglion


cells


using

NDIC


cresyl

optics


violet


staining


on unstained


and conventional


tissue


is that


optics.


the microscopist


The advantage


can focus


and down,


often


through


many


tissue


layers


(Figs


15-18),


and visualize


the three-dimensional


surfaces


of cells


processes


at successive


levels


without


identifying


eel


significant

1 types, a


light


benefit


attenuation.


that


This


unfortunately


greatly

does n


assists


ot translate


as well


into


a static,


two-dimensional


micrograph.


DMSO-clearing


yielded


particularly


good


results


with


the shark


pho-


toreceptor


layer,


the architecture


of which


can be otherwise


susceptible


to the shrinkage


effects


of xylene-clearing


and alcohol-dehydration.


Occasionally,


at the edges


of the retinal


wholemounts,


a teased-out


grouping


of neighboring


photoreceptors was


visible


in isolation


16).


Cones


were


readily


identifiable


in such


side


views


on the basis


of size,


shape,


and position


in the photoreceptor


layer.


Their


length


from


outer


limiting membrane


to the tip


of the


outer


segment


was 20


the widest


diameter


of the


cone


inner


segment


was about


consistent


with


Cohen


s (1980)


report


a cone


inner


segment width


of 5.1


juvenile


lemon


shark


retina.


Gruber


et al.


(1963)


reported


a cone


length























Fig.


Flatmount


contrast


(NDIC


view with
) optics


Nomarski


differential


of outer nuclear


layer


interference
containing t


nuclei
lemon
layer


of rod and


shark


located


cone


retina.


directly


photoreceptors


Bipolar


beneath


cell


in the wholemounted


nuclei


view in


of inner nuclear


White


bars


= 10 Um.






87












































I























Fig.


Isolated


grouping


of rod and


cone


photoreceptors


from


lemon


shark
outer
white


retina.


and inner


lines


mark


A single


segments


the level


cone


with


can be clearly


prominent


seen,


ellipsoid


of the rod ellipsoids


including
(arrow).


at which


counts


cone


density were


photoreceptor
(b) Focusing


nuclei
deeper,


the photoreceptor


made


in flatmount


lie just
a Muller


group


below


cell


seen.


view (see Fig. 17). The
outer limiting membrane.


process


NDIC


(arrow)


optics.


supporting


White


bars






89









SI
























Fig.


Photoreceptor mosaic


in flatmount


view


at level


of rod el-


lipsoids
central


and tips


retina


cone


of lemon


outer


shark.


segments


Many


cones


(see


Fig.


16a)


are present;


from
only


are indicated


(arrows)


Note


apparent


hexagonal


array


six rods
shark's


surrounding


visual


each


streak.


cone.


Same


This
view


area


lies


outside


within


of the visual


streak


in peripheral


retina.


Fewer


cones


can be


seen;


one is


indicated


(arrow).


NDIC


optics.


White


bars


= 10






91








U























Fig.


Ganglion


lemon
cells
fibers


shark re
can be s
running


cells


tina.
een s


from peripheral


in flatmount


A cluster


surrounded


toward


retina.


the optic


Note


view


from


of small
parallel


disc.


large


visual


and giant
bundles


Ganglion
process


axonal


streak of
ganglion


nerve
cells


(arrow)


from


giant


ganglion


cell.


NDIC optics.


White


bars


= 10





93