Exp Brain Res (1998) 120:291±305 � Springer-Verlag 1998

R E S E A R C H A R T I C L E

L. Herrero ´ F. Rodríguez ´ C. Salas ´ B. Torres

Tail and eye movements evoked by electrical microstimulationof the optic tectum in goldfish

Received: 24 September 1997 / Accepted: 4 December 1997

L. Herrero ´ B. Torres ())Lab. Neurobiología, Dept. Fisiología y Biología Animal,Fac. Biología, Univ. Sevilla, Avda. Reina Mercedes, 6,E-41012 Seville, Spaine-mail: btorres@cica.es, Fax: +34-5-4233480

F. Rodríguez ´ C. SalasLab. Psicobiología, Dept. Psicología Experimental, Univ. Sevilla,Seville, Spain

Abstract This work studies the tail and eye co-ordinatedmovements evoked by the focal electrical stimulation ofthe tectum in goldfish. The aim of the study is to under-stand better those tectal sites and mechanisms that eitherremain functionally unaltered or are adaptively modifiedacross vertebrates. Stimulation was applied in various tec-tal zones, and the characteristics of evoked tail and eyemovements were examined as a function of the stimula-tion site over tectal surface and the stimulus parameters.Two types of response were electrically evoked: the for-mer turned the body and the eyes contraversively towardsthe source of natural stimulus; the second produced initialipsiversive turning of the body and eyes, followed by sev-eral tail beats. Evoking one or other response depended onboth the site and parameters of stimulation, and responseswere interpreted as orienting- and escape-like, respective-ly. Depending on the stimulation site, four different zonesin the tectum were distinguished: in the medial zone thestimulus elicited eye and tail movements whose size in-creased with the distance to the rostral pole. The stimula-tion of the antero-medial zone evoked contraversive or ip-siversive eye saccades but tail movements were similar,irrespective of eye movements. Stimulation within the ex-treme antero-medial zone evoked convergent eye move-ments, and tail displacements turning the body either ip-siversively or contraversively. Stimulation of the posteri-or zone often evoked complex tail movements and purehorizontal eye saccades. Both orienting- and escape-likeresponses were also dependent on the stimulus parame-ters. The relationships between stimulus parameters andtail- and eye-orienting movement characteristics suggestthat the velocity and duration might be encoded in differ-ent aspects of the tectal activity. Current strength also

modified the number of tail beats that appeared during es-cape-like response. In conclusion, the present data suggestthe involvement of the optic tectum not only in orientingbut also in escape responses and that movements of eyeand tail mediating such responses depend on the tectal ac-tive locus together with its level of activity.

Key words Superior colliculus ´ Orienting ´ Escape ´Tecto-reticular pathway ´ Fish

Introduction

It is widely accept that the optic tectum (or the mamma-lian superior colliculus) plays an important role in thegeneration of the orienting movements towards novelsensory stimuli in vertebrates. Although the tectal mech-anisms involved in the generation of co-ordinated eye-head or -body movements are not well known (Guitton1992; Masino 1992), available information suggests thatthe contribution of the tectum to head movements variesmarkedly between species, increasing in those whose oc-ulomotor range is more limited (Guitton 1992). Thus, inbirds and amphibians, such as barn owls and toads, stim-ulation of any site within the tectum evokes only headand body movements (Ewert 1984; du Lac and Knudsen1990), while in monkeys and cats, whose oculomotorranges are � 55� and � 25�, respectively, co-ordinatedeye-head shifts are elicited from a large area of the col-liculus (Roucoux et al. 1980; Cowie and Robinson1994; ParØ et al. 1994; Freedman et al. 1996). Therefore,the tectal contribution to head movements is, at least inpart, aimed at displacing the visual axis in order to scanthe visual field beyond the oculomotor range. Currentknowledge about the tectal contribution to the kineticfeatures of eye- and body-orienting movements in fishis rather scarce (Akert 1949; Meyer et al. 1970; Al-Akelet al. 1986). Goldfish have a small oculomotor range(� 15�), but, in contrast to other vertebrates, the tectalcommand should be translated to the spinal circuitry togenerate body movements, since head movements per

292

se do not exist. This fact enhances the interest of the pres-ent study, whose first goal is to compare the tectal codi-fication of eye and tail movements with respect to theeye-head shifts evoked in other vertebrates. This ap-proach should allow the extension of general principlesof the tectal codification of motor commands across ver-tebrates as well as obtaining insights into specific proper-ties for fishes.

Although early studies suggested that the colliculuscodes the movement features depending on the active lo-cus alone (Robinson 1972), current evidence suggests thatthe neural processing underlying the tectal contribution toeye and gaze movements is much more complex. For in-stance: (1) the amplitude and velocity of eye (Van Opstalet al. 1990; Salas et al. 1997), head (du Lac and Knudsen1990), body (Ewert 1984; King et al. 1991) and gaze(ParØ et al. 1994; Freedman et al. 1996) movements aredependent on both the active tectal locus and stimulus pa-rameters; (2) the kinetic properties of eye (McIlwain1988) and gaze movements (Pelisson et al. 1989;Segraves and Goldberg 1991; Freedman et al. 1996) de-pend on the initial position; (3) the activity profile ofthe collicular neurons suggests the presence of a localfeedback loop carrying information about current eye po-sition to the colliculus (Munoz et al. 1991; Waitzman etal. 1991; Munoz and Wurtz 1995); and (4) local pharma-cological treatment of the colliculus modifies the main se-quence of elicited eye movements (Hikosaka and Wurtz1985; Lee et al. 1988). Therefore, a further goal of thiswork was to study whether any of such complex processesare present in the tectum of fish. With this aim, we inves-tigated (1) the influence of the tectal locus and stimulusparameters on evoked movements, (2) the dependenceof tail movements on the initial eye position, and (3)the relationships in the kinetic properties of eye- andtail-elicited movements.

A range of evidence also suggests that the superiorcolliculus of rodents is involved not only in the genera-tion of orienting movements but also in defensive-like re-sponses such as avoidance or flight (Ellard and Goodale1986; Sahibzada et al. 1986; Dean et al. 1988; Northmoreet al. 1988). Dean and co-workers (1989) suggested thatin animals such as rodents, with eyes located laterally forpanoramic vision, poorly developed fovea, and many pre-dators, the superior colliculus can take ªintelligentº deci-sions, which involve the animal�s turning towards oraway from the visual stimulus source. This critical deci-sion seems to be dependent on what area of the tectum isstimulated, as well as the depth within the tectum atwhich the stimulus is applied (Sahibzada et al. 1986;Northmore et al. 1988). Like rodents, goldfish are later-al-eyed; their visual field spans over 280� and lacks atrue-fovea. Although the involvement of tectum in es-cape-like responses in fish has occasionally been report-ed (Meyer et al. 1970; Al-Akel et al. 1986), little isknown about the influence of stimulus parameters andtectal locus on generating these responses in this biolog-ical group. Therefore, this was another purpose of thepresent report.

Materials and methods

Animals and surgical preparation

Experiments were carried out on eight goldfish (Carassius auratus)6±8 cm long from snout to tail base, obtained from local suppliers.The animals were kept in aquaria at 20�C for at least 2 weeks beforeany experiment.

Under general anaesthesia (1:20 000 wt/vol, solution of tricainemethane sulphonate), each animal was clamped firmly between twoplastic pads inside a home-made Perspex water chamber. The mouthwas fitted to a plastic tube connected to a well-aerated water circuitpropelled by a pump to ensure a constant flow over the gills. Anaes-thetic concentration in the water circuit was maintained at a suitablelevel during surgical preparation of the animals, which includedopening a hole through the skull and implanting coils near to the tailbase and on the eyes. The surgery was begun by cutting away asmall area of the cranium, dorsal to the midbrain, and removingthe underlying fat gently to avoid bleeding, until the tectal lobeswere clearly visible. To ensure head stability during recording,two thin rigid bars clamped the temporal bones of the fish to the tankwalls.

Tail and eye movements were recorded using the search coiltechnique. A ten-turn coil made of enamel-insulated 50-mm copperwire of 5 mm diameter was sutured 1.5 cm rostral to the tail baseand a 60-turn coil made of enamel-insulated 25-mm copper wireof 2.3 mm of diameter was sutured to the upper scleral margin ofeach eye. The tank containing the prepared fish was placed withinthe magnetic field with the tail in the centre (Fig. 1). The diameterof the magnetic field coil was 30 cm. Tail movement was calibratedbending the tail at known angles of � 10�, 20�, 30�, 40� and 50�. Eyemovement calibrations were obtained by rotating the magnetic fieldcoils at known angles (� 5�, 10� and 15�) around the stationary eyecoil. Calibrations were used to adjust the gain of the signals appro-priately. The small translational movements of the tail failed to shiftthe tail coil out of the uniform portion of the magnetic fields. Oncethe surgical procedure was finished, the water of the tank was chan-ged ± usually three to four times every 10 min ± thereby removingthe anaesthetic. The animals recovered an alert state about 10±15 min after the first change of water, as shown by the re-establish-ment of the normal pattern of eye movements (Salas et al. 1992).The zero tail position was defined while the animal was anaestheti-sed and represents the position at which the tail was not bent withrespect to the body axis. When the experiments began, about 1 h af-ter surgery, the zero eye position in the orbit was defined as the mid-dle of the eye movement range after 15 min of spontaneous eyemovements. Repetition of this procedure at the end of the experi-ment showed that the zero position of the tail and the eye did notdrift throughout the session. The electrical stimulus and recordingsof tail and eye movements were stored on videotape (Neuro-Corder,Neuro Data Instruments).

Electrical microstimulation

Electrical stimulation was carried out with monopolar glass-insulat-ed stainless steel wire (25 mm diameter) microelectrodes. In the ab-sence of stereotaxic measurement, the length and width of the visi-ble right optic tectum were carefully noted for each fish. Differentpenetrations were made with the same microelectrode, which wasadvanced with a microdriver. The location of each stimulation sitein the antero-posterior and medio-lateral axis was noted with respectto the previous measurements of the tectal surface. Because one goalof this work was to study the dependence of the stimulus site on thesize of the horizontal component of eye and tail movements, thepenetrations were carried out at fixed distance from the rostral tectalpole. Usually, the penetrations were made at 10, 25, 50, 75 and 90%from the rostral pole (Fig. 3), and the antero-posterior tectal axismeasured about 2.3 mm (� 200 mm range), while the distance inthe medio-lateral tectal axis was always maintained at 50% of thevisible tectum. Occasionally, other penetrations were carried out

293

at distances between those fixed, to determine the extension of thetectal functional zones (see Fig. 4). The microelectrode tip wasplaced between 300 and 400 mm in depth, at the level of the stratumgriseum centrale or the stratum album centrale at which the mini-mum current intensity (threshold range 10±30 mA) was needed toevoke tail and eye movements (Salas et al. 1997). To identify the lo-cation of the electrode tip, a small lesion was made using a singlecathodal pulse of 6 mA lasting 15 s. All data in this paper are fromstimulation sites on the right optic tectum.

Tail and eye movements were evoked with the above-mentionedmonopolar microelectrodes, using a train of cathodal pulses deliv-ered through a constant-current stimulus isolation unit. Usually theduration, pulse rate and pulse width of the stimulus train were fixedat 50 ms, 500 Hz and 0.2 ms; the current strength was increased insteps of 5 mA until obtaining the threshold (I) for movements (10±30 mA) and was delivered at 1.25 � I or 80 mA. When the experi-ments were to determine the effects of stimulus parameters, currentstrength (10±100 mA), pulse rate (50±500 Hz) and train duration(10±400 ms) were systematically varied. A group of at least teneye and tail movements was elicited to study the influence of thetectal locus and the stimulus parameters. Current strength was calcu-lated by monitoring the voltage across a 10-kW resistance in serieswith the stimulating electrode. A recovery time of 5 min was left be-tween penetrations; the experimental sessions lasted 3±4 h.

Histology

At the conclusion of the experimental session, the fish was re-ana-esthetised (1:5 000 solution of tricaine methane sulphonate) and per-fused transcardially with saline solution followed by 10% formalinin phosphate buffer. The brain was removed, frozen, and cut at50-mm sections that were stained with neutral red to locate the stim-ulation sites.

Data analysis

Videotape records were analysed on a processing digital-storage os-cilloscope (Tektronix, TDS 420). The horizontal component of elic-ited tail and eye movements, and their velocity, were displayed onthe oscilloscope screen. The following parameters were measured:initial and final position, amplitude, peak-velocity and latency ofhorizontal components of tail and eye movements. To representevoked saccades as vectors, the horizontal and vertical componentsof eye movements were transposed to an x-y co-ordinate system. Inthis system, 0� horizontal and vertical corresponded to the centre ofthe oculomotor range. To test the influence of both distance to therostral pole and stimulus parameters on the evoked movements, eachvalue of the stimulus parameters was applied ten times when the eyewas at the same position in the orbit, and when the tail was in itszero position. Figures 2±9 show the time course and trajectories ofevoked movements in the left eye and the tail following electricalstimulation of the right tectum.

Results

Classification of motor responses

The motor responses evoked by electrical microstimula-tion of the intermediate and deep tectal layers in goldfishwere classified into two groups. The first motor responsewas obtained at low current (10±100 mA) from the wholetectum. This motor response consisted of short-latency(30±60 ms) contraversive saccades of both eyes ± whichthen remained in a position close to the final after themovement (usually the evoked eye saccade was followedby an overshoot) for several hundred milliseconds, exceptwhen eye movements were evoked by the stimulation ofthe extreme antero-medial zone ± and a twitch of the con-tralateral axial muscles that induced a single tail beat(Fig. 2A). The tail displacement started several millisec-onds later than eye movement (10±20 ms), and, in con-trast to the eyes, the tail always re-centred after about150±250 ms. It must be added that, on some occasions,stimulation of the posterior zone (even using low current),failed to generate this motor pattern. Although in the pres-ent experiments the animals were not free to move, fromkinematical studies (Weihs 1973; Webb 1978; Eaton et al.1988; Domenici and Blake 1991), the sense of the singletail displacement implies body-turning contraversive tothe stimulated tectum. As discussed in detail later, we in-terpreted these co-ordinated movements of tail and eyesas an orienting-like response.

A second type of motor response was always obtainedwhen the stimulation was delivered at high currentstrength (more than 150 mA) from anywhere in the tectumand, usually, at low current strength (below 100 mA) fromsites within the caudal tectum. For this motor response,the elicited eye and tail movements were more complex(Fig. 2B). Thus, the stimulus evoked a first ipsiversiveeye movement of small amplitude (smaller than those ob-tained at low current strength), but the eyes, instead of re-maining close to their final positions after the movement,returned to a position near the centre of the orbit. Thisfirst eye movement was followed (or not), depending onthe presence (or absence) of tail beats, by several slow,

Fig. 1 Preparation for recording eye and tail movements in the alertgoldfish using the search coil technique. Body turns were restrictedby two plastic pads, whereas the tail was free to move (curved ar-row). The location of coils on the scleral and tail base, the microdriv-er for advancing the stimulating (St) glass-insulated microelectrodeand the Perspex-made fish chamber inside the coil frame are shown

294

small eye movements. The stimulus also elicited one orseveral tail beats, but, contrasting with the orienting re-sponse, the first tail movement was always mediated bythe activation of the ipsilateral axial muscles. In agree-ment with kinematical studies (Weihs 1973; Webb1978; Eaton et al. 1988; Domenici and Blake 1991), thistail displacement implies a body-turning ipsiversive to thestimulated tectum. As discussed later, these movementsresemble the natural escape response. The kinetic proper-ties of these responses, orienting and escape, depended ona combination of both the tectal stimulated locus andstimulus parameters.

Dependence of evoked eye and tail movement featureson the tectal stimulated locus

Low-current stimulation elicited tail and eye movementswhose amplitudes increased with the distance to the ros-tral pole (Fig. 3A). These increases in the amplitude werewell fitted to linear regressions for eye and tail (range ofcorrelation coefficients 0.58±0.85 and 0.54±0.95, respec-tively), for all four animals in which they were deter-mined. The slopes of these relationships ranged between1.86 and 3.64 for eye and 4.78 and 26.56 for tail move-ments, respectively (Fig. 3B). To build these relation-ships, only those tectal sites whose stimulation evoked asingle tail movement in the sense of the orientation weretaken into account, meaning that most of the results ob-tained after the stimulation of the posterior zone werenot included because they may not have correspondedto orienting movements (see Figs. 3Ac, 5).

As previously reported (Salas et al. 1994, 1997), theinfluence of the initial eye position on the elicited sac-cades at low current varied with the stimulated site, allow-ing four tectal zones to be distinguished (Figs. 4, 5). Thelargest tectal area was the medial zone. Stimulation deliv-ered in this zone evoked contraversive eye movementswith similar amplitudes and scarce influence of the initialeye position, and their trajectories were roughly parallel.The tail movements elicited from this tectal zone showed

Fig. 2A, B Time course of eye and tail movements elicited follow-ing the electrical stimulation of a site in the middle of the tectum(black dot in the drawing of the goldfish brain, insert at left) at80 mA (A) and 210 mA (B). Arrows show the direction of eye andtail movements. The drawings on the right show the idealised re-sponses of the animals in terms of contraversive (orienting-like)and ipsiversive (escape-like) body movements if they were free tomove. Note that in B the same eye and tail movements are represent-ed on two different time scales. Eh0, T0, Central positions for eyeand tail; _Eh0, _T0, zero in the velocity traces _Eh, _T , Eh, _Eh, T, _T trac-es of positions and velocity for eye and tail movements (Cb cerebel-lar body, OT optic tectum, TC tectal commissure, Tel telencephalon,VL vagal lobe)

295

very similar amplitudes and velocities (Fig. 4A). Electri-cal stimulation of the antero-medial tectal zone evokedgoal-directed eye movements and, so, contraversive andipsiversive saccades were evoked depending on the initialposition of the eye in the orbit. Tail-elicited movementsshowed very similar amplitudes and velocities, irrespec-tive of the amplitude and direction of the coupled eye sac-cades (Fig. 4B). Stimulation of the extreme antero-medialtectal zone evoked convergent eye saccades that droveboth eyes nasally to the edge of the orbit. Elicited tailmovements showed similar amplitudes and velocities, ir-

respective of saccadic amplitude, but in some animals(60%) the tail bent to evoke contraversive turning of thebody (Fig. 4C), while in other animals (40%) the tailmovement was to evoke ipsiversive turning. These lattermovements showed similar velocities to those that wouldevoke contraversive body-turning from this zone (bothabout 1000�/s), but smaller than those obtained whenthe movements resembled escape responses (over2000�/s).

The fourth tectal region is the posterior zone. As al-ready reported (Salas et al. 1994, 1997), stimulation with-

Fig. 3A, B Changes in ampli-tude and velocity across the an-tero-posterior tectal axis of eyeand tail movements evoked at1.25 � threshold. A Examples ofeye and tail movements evokedfrom three sites within differentzones (a, b, c), which are locat-ed on a dorsal and a parasagittalview of the tectum, obtained in asingle representative goldfish.Arrowhead denotes the startingof stimulation (st). B Left Linearrelationships of the variations inthe horizontal component of eyesaccades and tail movementswith the distance (DIST.) of thestimulating electrode to the ros-tral pole. Open circles and tri-angles indicate the amplitude oftail and eye movements ob-tained after each stimulation indifferent sites for the represen-tative animal illustrated in A.Right Similar linear relation-ships obtained for four differentanimals (VCb valvula cerebelli)

296

in this zone evoked contraversive eye movements whoseamplitudes and velocities depended on the initial eye po-sition. The most notable characteristic of these eye move-ments was the scarce, if present, vertical component,even when they were elicited from tectal sites near themidline (Fig. 5A). Three different patterns of tail move-ment were obtained from this posterior zone followingelectrical microstimulation at low current strength. Inthe first, the sense of tail bending evoked a contraversive

body-turning, and showed a similar time course to thatfound for the other tectal zones, such as the medial. Thispattern of tail movement was obtained in 25% of thestudied animals (Fig. 5B). The second pattern was morecomplex: a contraversive eye-turning coupled with sever-al tail beats. The first tail movement was of relatively lowamplitude and velocity (30� of amplitude and 1520�/s ofvelocity, for the case illustrated in Fig. 5C), at least com-pared with the second, which was larger (85�) and faster

Fig. 4A±C Eye and tail move-ments evoked at 80 mA fromsites within different tectalzones distinguished in this workand represented in the insert atthe top. A±C For a representa-tive single goldfish, the influ-ence of the initial eye positionon the amplitude, direction andvelocity of three eye and tailmovements (1,2,3) evoked fromthe same site within the medialzone (A), the antero-medialzone (B) and the extreme an-tero-medial zone (C). In thedrawings of fish, the arrows in-dicate the size and direction ofsuch movements. Right Thethree eye movements (1,2,3) interms of vectors, in which thevertical component of saccadeswas included. In C, the trajec-tories of saccades 1 and 2 aresuperimposed and arrowheadsindicate the starting eye positionfor these movements. At theright is also shown, for eachzone, the influence of initial eyeposition on the amplitude of eyemovements by linear relation-ships (C contraversive, I ipsi-versive, U upward, D down-ward)

297

(3500�/s). This second tail movement was of similar ve-locity (over 2000�/s) to those obtained after a high-cur-rent stimulus and would have evoked an ipsiversivebody-turn if the animal had been free to move. Then ap-peared several tail beats (never more than seven) whoseamplitudes and velocities progressively decreased. Thistype of complex tail movement appeared in 37.5% ofthe animals (Fig. 5C). The third pattern of tail movement(Fig. 5D) showed multiple tail beats (never less than six).The first one had an amplitude and velocity of similarvalue and time course to those observed when high cur-rent strength was delivered (see also Fig. 9), and the

sense of tail movement would yield body-turning ipsi-versive to the stimulated tectum. After this first tail turnappeared several beats whose amplitudes and velocitiesdecreased progressively, but the number, amplitudesand velocities were higher than those of the second pat-tern. Each animal showed only one of these three typesof tail response.

Dependence of tail and eye orienting-like movementson stimulus parameters

The effects of stimulus parameters on the metrics of eye-and tail-elicited movements were qualitatively observedat sites of the different tectal zones, except the posteriorzone, and quantitatively measured in four goldfish at thesame relative location sites within the medial zone. Atany site, the effects on eye and tail displacements ofchanges across a range of current strength and pulse ratewere similar, evoking mainly modifications in the ampli-tude and velocity of the movements. Figure 6 illustrates arepresentative example of the quantitative effects of mod-ifications in the current strength and pulse rate on themovement metrics obtained for a site located in the mid-dle of the visible tectum. Thus, the increase in current

Fig. 5A±D Eye and tail movements evoked at 80 mA from siteswithin the posterior zone. Insert at top left The location of the stim-ulating sites (open circle) on a dorsal view of the tectum. A Repre-sentative characteristics of the eye movements evoked from the pos-terior zone. From left to right: the time course of the horizontal (Eh)and vertical (Ev) components of two evoked saccades from differentinitial eye positions; velocity traces of these two saccades; trajecto-ries of the movements; and a linear relationship to show the influenceof the initial eye position on the amplitude of the evoked saccades.B±D The time course of eye (Eh) and tail (T) movements, as wellas their velocity traces ( _T), obtained for three representative goldfishafter the stimulation of sites within the posterior zone. In the draw-ings, the arrows indicate the direction and amplitude of eye and tailmovements

298

strength from 10 (usually the threshold) to 100 mA, main-taining the pulse rate (500 Hz) and train duration (50 ms)unaltered, caused an increase in the amplitude, velocityand duration of tail and eye movements (Fig. 6A). How-ever, once the current strength exceeded 80 mA, the am-plitude and velocity reached a saturating plateau for eyemovements, whereas some small increases were foundfor tail-evoked movements. When pulse rate was in-creased from 50 to 400 Hz, maintaining the currentstrength (80 mA) and train duration (50 ms) unchanged,the amplitude, velocity and duration of evoked eye and

tail movements also increased. Above 400 Hz, thesemovement metrics reached a saturating plateau (Fig. 6B).The increases in amplitude, velocity and duration of eyeand tail movements with the increase in current strengthand pulse rate were fitted by linear regressions for all fourgoldfish in which the effects of stimulus parameters onmovement features were quantitatively studied (range ofcorrelation coefficients 0.79±0.98). Comparing each pairof variables, the slopes of these linear regressions lines(range 0.02±16.6) were not different between fish (Stu-dent t-test, P>0.05).

Fig. 6A±C Effects of currentstrength and pulse rate on eyeand tail movement characteris-tics evoked from a single sitewithin the medial zone. A, B Atthe top is shown how the size ofeye and tail movements in-creases with current strength (A)and pulse rate (B). Other plotsindicate the effects of variationsin current strength and pulse rateon the mean values of ampli-tude, peak-velocity and durationof eye and tail movements.Vertical lines indicated SEM. CLinear relationships to show thecovariation of amplitude, peak-velocity and duration of eye andtail movements when these pa-rameters were modified by theeffect of current strength andpulse rate

299

Using these variations in stimulus parameters, it wasalso found that changes in the amplitude of evoked eyeand tail movements were related to modifications in thevelocity and duration of the movements (not illustrated),showing the features of the main sequences reported ingoldfish (Salas et al. 1997) and in other vertebrates (Fuchs1967); this suggests that, albeit with certain technical lim-itations, electrical stimulation of the tectum mimics theneural events mediating natural orienting response (duLac and Knudsen 1990; Cowie and Robinson 1994;Freedman et al. 1996; Salas et al. 1997). The amplitude,velocity and duration of eye and tail movements obtainedwith different current strength and pulse rate were plottedto study whether the modifications in the metrics of eye

and tail movements caused by changes in stimulus param-eters were inter-related. It was found that the movementmetrics changed proportionally, following well-fitted lin-ear relationships (range of correlation coefficients 0.81±0.98; Fig. 6C).

Modifications in stimulus train duration changed theamplitude and duration of eye and tail movements, but,in contrast to the main sequence, the velocity of suchmovements reached a saturating plateau earlier than am-plitude and duration. Thus, when the stimulus train dura-tion was increased from 10 to 100 ms, maintaining thecurrent strength (80 mA) and pulse rate (500 Hz) unal-tered, the amplitude and duration of eye and tail move-ments also increased, while above 40 ms the velocityreached a saturating plateau level (Fig. 7). The increasesin amplitude and duration of tail and eye movements withthe increase in stimulus train duration (10±100 ms) werefitted by linear regressions for all four studied animals(range of correlation coefficients 0.76±0.98). Comparingeach pair of variables, the slopes of these linear regres-sions lines (range 0.09±0.28) were not different between

Fig. 7 Effect of stimulus duration on eye and tail movement charac-teristics evoked from the same single site of Fig. 6 within the medialzone. At the top is shown how eye and tail movement amplitudes in-crease with stimulus duration. Other plots illustrate the effects ofvariation in stimulus train duration on the mean values of amplitude,peak-velocity and duration of eye and tail movements. Vertical linesindicated SEM. Arrowhead denotes the starting of stimulation (st)

Fig. 8 Effect of stimulus trains lasting more than 100 ms on thetime course of eye and tail movements evoked from a single sitewithin the medial zone. Note that stimuli lasting 200 ms and400 ms evoked several tail movements and some of them were un-coupled with eye movements (arrows). Arrowhead denotes the start-ing of stimulation (st)

300

fish (Student t-test, P>0.05). Finally, a notable effect wasoccasionally found (in two of four animals) when thestimulus train lasted 200 ms or more. In this case, thestimulus evoked one or two (staircase-linked) eye sac-cades that maintained their final position after the move-ment, while the tail beat several times depending on howlong the stimulation lasted (Fig. 8).

Dependence of escape-like movementson stimulus parameters

As already described (see classification of the motor re-sponses; Fig. 2B), stimulation of the tectum at high cur-rent (over 150 mA) generated eye movements and a vari-able number of tail displacements. The first tail move-ment showed a very high peak velocity (always over

2000�/s) and the sense of the displacement would haveturned the animal ipsiversively to the stimulated tectum.We have classified this motor response as an escape-likeresponse, and it was found in all animals studied (n=8),from any tectal site.

The features of these tail movements were always de-pendent on the stimulus parameters. Figure 9 shows thetime course of tail movements and the quantitative dataobtained after delivery of different current strengths to asite of the medial zone for a single representative gold-fish. Thus, stimulation from 150 to 400 mA caused an in-crease in the number of tail beats from 1 to 15 in the 5 sfollowing the stimulus (Fig. 9A). Furthermore, stimula-tion with different intensities showed that the number oftail beats depended on the current strength delivered(Fig. 9B) and that the number of tail beats after the stim-ulus decayed with time (Fig. 9C).

Fig. 9A±C Effects of currentstrength parameters on escape-like tail movements evokedfrom a site within the medialzone in a single representativegoldfish. A Time course of tailmovements following stimula-tion at different current strength.B Histogram showing the influ-ence of current strength on thetotal number of tail beats ob-tained in the five seconds afterstimulation. C Total number oftail beats in each second afterstimulation depending on cur-rent strength. Vertical lines in-dicate SEM

301

Discussion

Contraversive vs. ipsiversive body-turn

Depending on the features of tail and eye movementsevoked by tectal electrical microstimulation, two groupsof motor responses have been reported here. A stimulusdelivered with low current strength (less than 100 mA)within the third or fourth tectal rostral sites, and occasion-ally from the whole tectum, evoked a tail beat by the con-traction of the contralateral axial muscles, which wouldgive a contraversive body-turn if the goldfish were unre-strained (Weihs 1973; Webb 1978; Eaton et al. 1988;Domenici and Blake 1991). Depending on the tectal site,this low-current stimulation also evoked eye movementswhose direction and amplitude were roughly aligned withthe retinotopic visual arrangement within overlying tectallayers (as we discussed in Salas et al. 1997). Taking bodyand eye movements together, we classified this responseas orienting-like, because it generates an eye and bodyturn towards the potential source of sensorial stimulusin space. In fact, these tail and eye movements resemblethose of natural orienting gaze movements in goldfish(Herman and Constantine 1971; Easter et al. 1974) andalso in other vertebrates, since they are mediated by co-ordinated eye and head movements, which involve thecontraction of the neck muscle contralateral to the activecolliculus and a contraversive eye saccade (Bizzi et al.1971; Roucoux et al. 1980; Roucoux and Crommelinck1988). Evidence that the tectum or the superior colliculuscodes orienting movements has been extensively reportedin different species of fish (Akert 1949; Meyer et al. 1970;Al-Akel et al. 1986), amphibians (Ingle 1983; Ewert1984; Masino and Grobstein 1989a, b), reptile (Schapiroand Goodman 1969; Stein and Gaither 1981), bird (duLac and Knudsen 1990), and mammals (Schaefer 1970;Roucoux et al. 1980; Sahibzada et al. 1986; Northmoreet al. 1988; Segraves and Goldberg 1991; ParØ et al.1994; Freedman et al. 1996).

Electrical stimulation within intermediate and deeptectal layers with high current strength (more than150 mA), and occasionally with current strength below100 mA from the posterior tectum, initially evoked thecontraction of the ipsilateral axial muscles and a tail-turnthat would give an ipsiversive body-turn if the animalwere free to move (Weihs 1973; Webb 1978; Eaton etal. 1988; Domenici and Blake 1991), and ipsiversiveeye movements, as opposed to orienting movements. Be-cause this response would move the animal away from thepotential source of natural stimulus, we classified it as anescape-like response (Dean et al. 1989). The tectal in-volvement in the generation of escape responses was de-scribed early in fish (Akert 1949; Fiedler 1968; Meyer etal. 1970; Al-Akel et al. 1986) and has also been reportedin amphibians (Ingle 1983; Ewert 1984; Masino andGrobstein 1989a, b; Roche King and Comer 1996) and ro-dents (Ellard and Goodale 1986; Sahibzada et al. 1986;Northmore et al. 1988). Dean and co-workers (1989) sug-gested that the rodent�s superior colliculus can trigger the

decision of orienting versus flight, depending on the fea-tures of the stimulus and on its location in the visual field(Sahibzada et al. 1986; Dean et al. 1989). This view of thecollicular functional role is based on an ethological inter-pretation of benefits and cost that involve moving towardsor away from the stimulus (Dean et al. 1989).

The two evoked responses reported here depended on acombination of both the tectal stimulated site and currentstrength. These results might also be interpreted followingthe ethological view posited by Dean et al. (1989). In fact,the importance of the stimulus characteristics has been re-ported as a critical factor to evoke orienting or avoidanceresponses in both amphibians (Ewert 1984) and fish(Eaton et al. 1988). Furthermore, if a stimulus appearsfrom behind the animal, activating the posterior tectum,it is treated as potentially dangerous, and so it evokespreferentially an escape response. In the following para-graphs we will discuss the influence of the tectal siteand stimulus parameters on the features of both orienting-and escape-like responses.

Dependence of tail and eye orienting- and escape-likemovements on the tectal stimulated site

When a low-current stimulus was applied along the an-tero-posterior axis of most of the tectum (except the pos-terior zone), and occasionally for the whole tectum, theamplitude of eye and tail movements increased. This re-sult suggests the alignment of the visual and motor mapsin goldfish, since, when the stimulation site was locatedmore caudally within the tectum, the amplitude of thebody and eye movements increased, orienting the animaltowards more peripheral zones of the visual field. Thealignment of the sensorial and motor maps within the tec-tum is a common characteristic of all vertebrate speciesstudied (Masino 1992), and it has been suggested as anessential element for the circuitry that allows the sensori-motor transformation to perform orienting movements(Robinson 1972; Munoz and Wurtz 1995).

As already stated in the Introduction, comparativestudies suggest that the tectal contribution to eye and headmovements is species-specific. Thus, the tectal contribu-tion to head displacement rises for species with smallestoculomotor range, when head movement becomes neces-sary to reach targets beyond oculomotor range (Guitton1992). From this point of view and taking into accountthat the motility of the eye in goldfish is not more than� 15� (Easter 1971; Herman and Constantine 1971), itshould not be surprising that stimulation of the tectal sitesclose to the rostral pole evokes body movements; in otherwords, the whole tectum evokes body (besides eye) move-ments. In contrast, head movements in cats and monkeysare evoked only from the caudal zones of superior collic-ulus (Roucoux et al. 1980; Segraves and Goldberg 1991;ParØ et al. 1994; Cowie and Robinson 1994), whereas inrodents (Northmore et al. 1988), barn owls (du Lac andKnudsen 1990) and amphibians (Ewert 1984) they arealso evoked from the whole tectum.

302

Different functional zones in the goldfish tectum

On the basis of the influence of the initial eye position onthe characteristics of evoked eye movements, we havepreviously distinguished four different zones in the gold-fish tectum, which could subserve adaptive species differ-ences in the functional role of this structure comparedwith other vertebrates (Salas et al. 1994, 1997).

The medial zone is the biggest area in the tectum; stim-ulation delivered in this zone evoked eye movementswhose direction changed from upward to downward whenthe stimulation site was displaced in the mediolateral axis,as well as eye saccades whose horizontal component in-creased with the distance of the stimulus to the rostral poleof the tectum (Salas et al. 1997). These results, togetherwith those reported here of increased size and velocity oftail displacements across the tectal antero-posterior axis,and others obtained following tectal electrical stimulationin free-swimming fish, in which the animal performed turnsin different directions depending on the stimulated site(Fiedler 1968; Meyer et al. 1970; Al-Akel et al. 1986), sug-gest that this zone is involved in triggering orienting re-sponses towards different sites of the surrounding space.

The antero-medial zone is characterised by the pres-ence of goal-directed eye movements. Thus, stimulationdelivered in this zone evoked ipsiversive and contraver-sive eye saccades of different size and velocity dependingon the initial eye position, whereas, irrespective of thesefeatures of the eye movements, the tail displacementswere similar in both size and velocity (see Fig. 4B).The uncoupling between eye and tail movement featuresis a quite remarkable phenomenon, because it gives in-sights into the mechanism of tectal codification of eyeand body movements in fish that might perhaps be ex-tended to eye-head gaze shifts in other vertebrates. Amodel has recently been proposed in which the tectumis inside a local feedback loop that compares current gazewith the desired gaze movement, and the output is a singlesignal of gaze motor error that drives eye and head motorcircuits in the brainstem (Guitton 1992). The presence ofa feedback circuit for gaze control is supported by exper-iments in which perturbations in the signal received bythe comparator ± current gaze ± are compensated bychanges in the amplitude of gaze displacement (Pelissonet al. 1989, 1995). However, these studies of gaze pertur-bation fail to explain the influence of the variations onsolely one element of gaze (eye, head or body) in the per-formance of the co-ordinated movements of gaze shift. Inthis work, we have the opportunity to observe the influ-ence of initial eye position on tectal-evoked eye-tailmovements. The results indicate that, while eye move-ments are strongly influenced by initial eye position lead-ing to either ipsiversive or contraversive eye movements,the tail movement remains unaffected by initial positionof the eye. The present results seem to be inconsistentwith those showing that head contribution to gaze shiftsdepends on initial position of the eye in the orbit (Freed-man et al. 1996; Freedman and Sparks 1997). This dis-crepancy could be attributable to differences in the exper-

imental preparation, since in the present study, headmovements were restrained, and so the signal of currenthead displacement provided by the vestibular system dur-ing gaze shifts is not feeding the comparator of the feed-back circuitry to achieve the orienting movement. Theseresults could suggest that although the neural processingof eye and body signals can converge in some places ofthe feedback circuits, as shown by the perturbation exper-iments (Pelisson et al. 1989, 1995), they can also be pro-cessed separately by the gaze system, in such a way thatmodification in one of the systems could (Freedman et al.1996; Freedman and Sparks 1997) or could not (as shownhere) involve changes in the other.

Another feature of the proposed feedback model is thatthe superior colliculus generates a single signal related togaze shifts driving both head and eye movements (Munozet al. 1991; Guitton 1992; Freedman et al. 1996). Never-theless, some of the present results cast doubt on this hy-pothesis in fish. Thus, when a stimulus train lasting be-tween 200 and 400 ms was delivered, the eye remainedstationary in its final position while the tail beat repeti-tively, which implies that, following tectal electrical stim-ulation, tail movements can appear either combined withor isolated from eye movements. Furthermore, the corre-lation coefficients of the linear regressions (not shown)between the latencies of tail and eye movements and be-tween amplitudes or velocities or durations of such dis-placements were usually low (r<0.5). On the basis ofsimilar findings obtained following the electrical stimula-tion of monkey superior colliculus, it has been suggestedthat the colliculus has at least two output systems control-ling eye and head motor centres (Cowie and Robinson1994). In goldfish, the evidence suggests that either thereare separate tectal commands for body and eye motor cen-tres in the brainstem (Cowie and Robinson 1994), or eachone of these motor systems interprets the same commandin accord with its own peculiar functional characteristics(Masino 1992; Freedman et al. 1996).

Stimulation delivered with current strength below100 mA within the extreme antero-medial zone evokedconvergent eye movements and a tail beat of smaller sizeand velocity to that evoked with the same current strengthfrom the medial and posterior zones, which would let theanimal turn either ipsiversively or contraversively. Uni-tary recordings of sensory cells within this extremeantero-medial tectal zone show a representation of the vi-sual fields extending a few degrees within ipsilateral andcontralateral nasal sides (Jacobson and Gaze 1964;Schwassmann and Kruger 1965; Trevarthen 1968). Sucha visual representation might be underlying both ipsi-versive and contraversive body-turns and convergenteye movements evoked by the stimulation of the interme-diate and deep layers of this tectal zone (Salas et al.1997). These movements probably allow the visual stim-ulus to be projected on both central retinas and could bepart of highly adaptive tectal mechanisms for lateral-eyedfishes aimed at catching food with accuracy (Salas et al.1997). The involvement of this tectal zone in feeding re-sponses is also supported by complementary data in free-

303

swimming fish. Thus, the animal performs forward move-ments in a ªfood searchingº manner following the electri-cal stimulation of anterior tectal zones (Fiedler 1968;Meyer et al. 1970; Al-Akel et al. 1986) and fails to pursueand catch food pellets after bilateral tectal ablation (Yageret al. 1977).

Stimulation of the posterior zone with current strengthbelow 100 mA evoked a contraversive movement of theeyes and (depending on the animal) a different patternof tail movement. A first type of tail movement (seeFig. 5B) might be classified as an orienting-like responsebecause it turns the animal towards the contralateral spaceand the velocity of this movement never rises above2000�/s. These tail movement features in fish are in agree-ment with the topography of the motor map reported inother vertebrates for head movement from this posteriorzone, since its stimulation evokes the fastest and largesthead movements (du Lac and Knudsen 1990; Cowieand Robinson 1994; ParØ et al. 1994).

Stimulation of the posterior zone can also evoke(Fig. 5C, D) another two types of fast tail movement thatinitially turn the body ipsiversively and continue with sev-eral tail beats that move the animal away from the poten-tially dangerous natural stimulus. On the basis of thevelocity and sense of tail displacement, this responsewas classified as escape-like. The presence of tectal zonesfrom which it is easier to obtain a defensive-like responsethan an orienting one has also reported in rodents(Sahibzada et al. 1986; Dean et al. 1988; Northmore et al.1988). For instance, chemical (Dean et al. 1988) or elec-trical (Sahibzada et al. 1986) stimuli delivered into the an-tero-medial superior colliculus of rats evoked differentbehaviour resembling natural defensive responses. Be-cause orienting and escape responses are mediated bythe tectal contralateral and the ipsilateral descending path-ways (Dean et al. 1986; Ellard and Goodale 1986; Westbyet al. 1990), and the cells of origin of these tracts are notthe same for both (Redgrave et al. 1986), the tectum hasbeen proposed as a functional mosaic (Dean et al. 1989)in which, following an ethological interpretation, areashave been adaptively specialised to produce avoidance re-sponses when the stimulus appears from above (as thepredator generally does), activating the antero-medial ar-ea of the tectum in rodents; while prey, which is found onthe ground, mainly activates the caudo-lateral tectum(Dean et al. 1989; Westby et al. 1990). Studies in free-swimming fish also suggest that the posterior tectum isan adaptively specialised area involved in flight reaction± thus the electrical stimulation of this zone evokes a sud-den halting of normal activity, dorsal fin erection, rapidturning and swimming to the edge of the tank (Meyer etal. 1970; Al-Akel et al. 1986).

Effects of stimulus parameters on tailand eye orienting movements

The present results show that the metric and kinetic prop-erties of tail and eye movements depend on both the tectal

locus stimulated and stimulus parameters. These data areincompatible with the suggestion of early reports that thetectal active site alone encodes movement features(Robinson 1972), but agree with more recent results show-ing a dependence of eye (Guitton et al. 1980; Van Opstal etal. 1990; Stanford et al. 1996), gaze (Segraves and Gold-berg 1991; ParØ et al. 1994; Freedman et al. 1996), head(du Lac and Knudsen 1990) and body (Northmore et al.1988; King et al. 1991) movement features on stimulus pa-rameters. Because changes in stimulus parameters influ-ence movement features through modifications in the fir-ing rate and number of collicular active cells (Yeomans1990), present data are also in accord with recent neuro-physiological data suggesting a dual (spatial plus temporal)codification of movements by the activity of neurons lyingin intermediate and deep tectal layers (Munoz et al. 1991;Waitzman et al. 1991; Munoz and Wurtz 1995), and theymight support the suggestion that these tectal mechanismsare largely conserved across vertebrate phylogeny.

The relationships between stimulus parameters and tailand eye movements suggest that velocity and duration ofmovements might be encoded in different aspects of tectalactivity. Thus, increases in current strength (10±120 mA)and pulse rate (25±400 Hz) lead to similar effects (in-creases in the velocity and size of movements), whereasincreases in train duration over 40 ms modify size and du-ration of movements without affecting their velocities.Similar results have been found when studying the effectof changes in electrical stimulus parameters on eye move-ments evoked from the superior colliculus in monkeys(Stanford et al. 1996) and the tectum in goldfish (Salaset al. 1997), and they have been interpreted as supportingthe suggestion that the size of movements can be modi-fied by separate signals, affecting either the velocity, bychanges in the total amount of tectal activity, or the dura-tion, by changes in the period of time in which a tectal lo-cus is active (Stanford et al. 1996; Salas et al. 1997). Theinfluence of the level of tectal activity on the velocity ofeye movements might also be inferred from studies inwhich a covariation of the activity profile of some colli-cular neurons with the instantaneous eye velocity duringsaccades was found (Berthoz et al. 1986; Munoz et al.1991). Furthermore, the presence of separate tectal mech-anisms to code the size of movements can also explainwhy some drugs modify the main sequence of eye move-ments, e.g. decreases in the peak-velocity of movementswithout proportional effects on amplitude (Hikosaka andWurtz 1985; Lee et al. 1988). The present results allowthe suggestion that such separate tectal signals can beused to code movement features not only in the premotorcentres for eye movements but also for body movements.

Influence of current strength on escape-like response

The present data show that, by stimulating from the samesite but increasing current strength over 150 mA, the di-rection of turning of the first tail movement changes.Thus, instead of a tail displacement of not more than

304

1500�/s that moves the body contraversively, there is afaster tail movement (over 2000�/s) that moves the bodyipsiversively. The change from orienting- to avoidance-like responses, as a consequence of an increase in currentstrength above a certain threshold, with stimulus deliv-ered within the tectum, has already been reported in ro-dents (Sahibzada et al. 1986; Northmore et al. 1988)and fish (Akert 1949). Akert (1949), using gross elec-trodes applied to the tectal surface of anaesthetised rain-bow trout (Salmo gairdnieri), reported that low- andhigh-current stimuli evoke antagonistic responses with re-gard to body bending, suggesting that systems for the twotypes of body turn are located at different depth within thetectum. A similar suggestion has also been made in rela-tion to rodents (Dean et al. 1986; Sahibzada et al. 1986;Northmore et al. 1988; Westby et al. 1990). Alternativeexplanations can be posed, such as that high-current stim-ulus evokes escape-like response by spread of current be-yond tectal layers (Yeomans 1990), in particular, to thereticulospinal system ± a neural system involved in escaperesponse in both fishes (Eaton et al. 1988; Nissanov andEaton 1989) and other vertebrates (Ingle 1983; Davis1984; Roche King and Comer 1996). Although this latterexplanation cannot be ruled out as a result of the presentexperiment, it seems much more difficult to explain whysome local injections of drugs within the tectum evokeavoidance response (Dean et al. 1988).

In addition, the fact that both orienting and escape re-actions can be elicited from the same tectal zones (onlyvarying the current strength) could indicate that both mo-tor reactions are potentially aligned with the sensory map.This may reveal that the escape reactions are also orientedwith reference to the location of the stimulus. This sug-gestion seems compatible both with behavioural data infish showing that, for the success of an escape response,it is determinant where the animal moves in relation tothe predator (Weihs and Webb 1984), and also with therecently proposed role for the reticulo-spinal system inthe repertoire of trajectories of escape response in teleosts(Eaton et al. 1988; Nissanov and Eaton 1989; Foremanand Eaton 1993).

The present data also show that high-current stimula-tion evokes not only a first tail movement that bendsthe animal in the opposite direction to the source of stim-ulus but also a variable number of tail beats. This chain ofmotor events could correspond with stages 1 and 2 of es-cape responses (Eaton et al. 1988). Thus, in the first stage,the body of the fish assumes a C-like shape, whose direc-tion is opposite to the source of the stimuli (Foreman andEaton 1993), while during the second stage the fish accel-erates away from the stimulus (Eaton et al. 1988); such apropulsive displacement of the goldfish could be mediat-ed by the tail beats initiated after tectal stimulation. Asshown here, current strength influences the number of tailbeats and thus the capacity of the goldfish to get far fromthe stimulus if the animal is free to swim. The variabilityin the propulsive phase has also been reported in free-swimming fish as a function of visual stimulus character-istics (Weihs and Webb 1984). Therefore, the amount of

tectal activity, which depends on current strength stimula-tion (present data), could be wired to the level of activityof the reticular formation that drives tail-beat frequency(Fetcho and Svoboda 1993).

Acknowledgements This work was supported by grants of DGI-CYT no. PB-93-0916 and PB-96-1334, Accion Integrada H-F, andthe Junta de Andalucía.

References

Akert K (1949) Der visuelle Greifreflex. Helv Physiol PharmacolActa 7:112±134

Al-Akel AS, Guthrie DM, Banks JR (1986) Motor responses to lo-calized electrical stimulation of the tectum in the freshwaterperch (Perca fluviatilis). Neuroscience 19:1381±1391

Berthoz A, Grantyn A, Droulez J (1986) Some collicular efferentneurons code saccadic eye velocity. Neurosci Lett 72:289±294

Bizzi E, Kalil RE, Tagliasco V (1971) Eye-head coordination inmonkeys: evidences for centrally patterned organization. Sci-ence 173:452±454

Cowie RJ, Robinson DL (1994) Subcortical contributions to headmovements in macaques. I. Contrasting effects of electricalstimulation of a medial pontomedullary region and the superiorcolliculus. J Neurophysiol 72:2648±2664

Davis M (1984) The mammalian startle response. In: Eaton RC (ed)Neural mechanism of startle behavior. Plenum, New York, 287±342

Dean P, Redgrave P, Sahibzada N, Tsuji K (1986) Head and bodymovements produced by electrical stimulation of superior collic-ulus in rats: effects of interruption of crossed tecto-reticulo-spi-nal pathway. Neuroscience 19:367±380

Dean P, Mitchell IJ, Redgrave P (1988) Responses resembling de-fensive behaviour produced by microinjection of glutamate intosuperior colliculus of rats. Neuroscience 24:501±510

Dean P, Redgrave P, Westby GWM (1989) Event or emergency?Two response systems in the mammalian superior colliculus.Trends Neurosci 12:137±147

Domenici P, Blake RW (1991) The kinematics and performance ofthe escape response in the angelfish (Pterophyllum eimekei). JExp Biol 156:187±204

Easter SS (1971) Spontaneous eye movements in restrained gold-fish. Vision Res 11:333±342

Easter SS, Johns PR, Heekenlively D (1974) Horizontal compensa-tory eye movements in goldfish (Carassius auratus). I. The nor-mal animal. J Comp Physiol 92:23±35

Eaton RC, DiDomenico R, Nissanov J (1988) Flexible body dynam-ics of the goldfish C-start: implications for reticulospinal com-mand mechanisms. J Neurosci 8:2758±2768

Ellard CG, Goodale MA (1986) The role of the predorsal bundle inhead and body movements elicited by electrical stimulation ofthe superior colliculus in the mongolian gerbil. Exp Brain Res64:421±433

Ewert JP (1984) Tectal mechanisms that underlie prey-catching andavoidance behavior in toads. In: Vanegas H (ed) Comparativeneurology of the optic tectum. Plenum, New York, 247±416

Fetcho JR, Svoboda KR (1993) Fictive swimming elicited by elec-trical stimulation of the midbrain in goldfish. J Neurophysiol70:765±780

Fiedler K (1968) Verhaltenswirksame strukturen im fischgehira.Verh Dtsch Ges Zool 31:602±616

Foreman MB, Eaton RC (1993) The direction change concept for re-ticulospinal control of goldfish escape. J Neurosci 13:4101±4113

Freedman EG, Sparks DL (1997) Eye-head coordination duringhead unrestrained gaze shifts in rhesus monkeys. J Neurophysiol77:2328±2348

Freedman EG, Stanford TR, Sparks DL (1996) Combined eye-headgaze shifts produced by electrical stimulation of the superiorcolliculus in rhesus monkeys. J Neurophysiol 76:927±952

305

Fuchs AF (1967) Saccadic and smooth pursuit eye movements in themonkey. J Physiol (Lond) 191:609±631

Guitton D (1992) Control of eye-head coordination during orientinggaze shifts. Trends Neurosci 15:174±179

Guitton D, Crommelinck M, Roucoux A (1980) Stimulation of thesuperior colliculus in the alert cat. I. Eye movements and neckEMG activity evoked when the head is restrained. Exp BrainRes 39:63±73

Herman HT, Constantine MM (1971) Eye movements in the gold-fish. Vision Res. 11:313±331

Hikosaka O, Wurtz RH (1985) Modification of saccadic eye move-ments by GABA-related substances. I. Effects of muscimol andbicuculine in monkey superior colliculus. J Neurophysiol 53:266±291

Ingle D (1983) Brain mechanisms of visual localization in frogs andtoads. In: Ewert JP, Capranica RR, Ingle D (eds) Advances invertebrate neuroethology. Plenum, New York, pp 177±226

Jacobson M, Gaze RM (1964) Types of visual response from singleunits in the optic tectum and optic nerve of the goldfish. Q J ExpPhysiol 49:199±209

King SM, Dean P, Redgrave P (1991) Bypassing the saccadic pulsegenerator: possible control of head movement trajectory by ratsuperior colliculus. Eur J Neurosci 3:790±801

Lac S du, Knudsen EI (1990) Neural maps of head movement vectorand speed in the optic tectum of the barn owl. J Neurophysiol63:131±146

Lee C, Rohrer WH, Sparks DL (1988) Population coding of saccadiceye movements by neurons in the superior colliculus. Nature332:357±360

Masino T (1992) Brainstem control of orienting movements: intrin-sic coordinate systems and underlying circuitry. Brain BehavEvol 40:98±111

Masino T, Grobstein P (1989a) The organization of descendingtectofugal pathways underlying orienting in the frog, Rana pipi-ens. I. Lateralization, parcellation and an intermediate spatialrepresentation. Exp Brain Res 75:227±244

Masino T, Grobstein P (1989b) The organization of descendingtectofugal pathways underlying orienting in the frog, Rana pipi-ens. II. Evidence for the involvement of a tecto-tegmental-spinalpathway. Exp Brain Res 75:245±264

McIlwain JT (1988) Effects of eye position on electrically evokedsaccades: a theoretical note. Vis Neurosci 1:239±244

Meyer DL, Schott D, Schaefer KP (1970) Reizversuche im tectumopticum freischwimmender Kabeljaue bzw. Dorsche (Gadusmorrhua L.). Pflugers Arch 314:240±252

Munoz DP, Wurtz RH (1995) Saccade related activity in monkey su-perior colliculus. II Spread of activity during saccades. J Neuro-physiol 73:1334±1348

Munoz DP, Guitton D, Pelisson D (1991) Control of orienting gazeshifts by the tectoreticulospinal system in the head free cat. III.Spatiotemporal characteristics of phasic motor discharges. JNeurophysiol 66:1642±1666

Nissanov J, Eaton RC (1989) Reticulospinal control of rapid escapeturning maneuvers in fishes. Am Zool 29:103±121

Northmore DPM, Levine ES, Schneider GE (1988) Behavior evokedby electrical stimulation of the hamster superior colliculus. ExpBrain Res 73:595±605

ParØ M, Crommelinck M, Guitton D (1994) Gaze shifts evoked bystimulation of the superior colliculus in the head-free cat con-form to the motor map but also depend on stimulus strengthand fixation activity. Exp Brain Res 101:123±139

Pelisson D, Guitton D, Munoz DP (1989) Compensatory eye andhead movements generated by the cat following stimulation-in-duced perturbations in gaze position. Exp Brain Res 78:654±658

Pelisson D, Guitton D, Goffart L (1995) On-line compensation ofgaze shifts perturbed by micro-stimulation of the superior collic-ulus in the cat with unrestrained head. Exp Brain Res 106:196±204

Redgrave P, Odenkunle A, Dean P (1986) Tectal cells of origin ofpredorsal bundle in rat: location and segregation from ipsilateraldescending pathway. Exp Brain Res 63:279±293

Robinson DA (1972) Eye movements evoked by collicular stimula-tion in the alert monkey. Vision Res 12:1795±1808

Roche King J, Comer CM (1996) Visually elicited turning behaviorin Rana pipiens: comparative organization and neural control ofescape and prey capture. J Comp Physiol 178:293±305

Roucoux A, Crommelinck M (1988) Control of head movement dur-ing visual orientation. In: Peterson BW, Richmond FJ (eds) Con-trol of head movement. Oxford University Press, New York, pp208±223

Roucoux A, Guitton D, Crommelinck M (1980) Stimulation of thesuperior colliculus in the alert cat. I. Eye and head movementsactivity evoked when the head is unrestrained. Exp Brain Res39:75±85

Sahibzada N, Dean P, Redgrave P (1986) Movements resemblingorientation or avoidance elicited by electrical stimulation ofthe superior colliculus in rats. J Neurosci 6:723±733

Salas C, Navarro F, Torres B, Delgado-García JM (1992) Effects ofdiazepam and D-amphetamine on rhythmic pattern of eye move-ment in goldfish. Neuroreport 3:131±134

Salas C, Herrero L, Rodriguez F, Torres B (1994) On the role ofgoldfish optic tectum in the generation of eye movements. In:Delgado-García JM, Godaux E, Vidal PP (eds) Information pro-cessing underlying gaze control. Elsevier, New York, pp 87±95

Salas C, Herrero L, Rodriguez F, Torres B (1997) Tectal codifica-tion of eye movements in goldfish studied by electrical micro-stimulation. Neuroscience 78:271±288

Schaefer KP (1970) Unit analysis and electrical stimulation in theoptic tectum of rabbits and cats. Brain Behav Evol 3:222±240

Schapiro H, Goodman DC (1969) Motor functions and their anatom-ical basis in the forebrain and tectum of the alligator. Exp Neu-rol 24:187±195

Schwassmann HO, Kruger L (1965) Organization of the visual pro-jections upon the optic tectum of some freshwater fish. J CompNeurol 124:113±126

Segraves MA, Goldberg ME (1991) Properties of eye and headmovement evoked by electrical stimulation of the monkey supe-rior colliculus. In: Berthoz A, Graf W, Vidal PP (eds) The head-neck sensory-motor system. Oxford University Press, NewYork, 292±295

Stanford TR, Freedman EG, Sparks DL (1996) The site and param-eters of microstimulation: evidence for independent effects onthe properties of saccades evoked from the primate superior col-liculus. J Neurophysiol 76:3360±3381

Stein SE, Gaither NS (1981) Sensory representation in reptilian op-tic tectum: some comparisons with mammals. J Comp Neurol202:69±87

Trevarthen CB (1968) Vision in fish: the origins of the visual framefor action in vertebrates. In: Ingle D (ed) The central nervoussystem and fish behavior. Chicago University Press, Chicago,pp 61±94

Van Opstal AJ, Van Gisbergen JAM, Smit AC (1990) Comparisonof saccades evoked by visual stimulation and collicular electri-cal stimulation in the alert monkey. Exp Brain Res 79:299±312

Waitzman DM, Ma TP, Optican LM, Wurtz RH (1991) Superiorcolliculus neurons mediate the dynamic characteristics of sac-cades. J Neurophysiol 66:1716±1737

Webb PW (1978) Fast start performance and body form in sevenspecies of teleost fish. J Exp Biol 47:211±226

Weihs D (1973) The mechanism of rapid starting of a slender fish.Biorheology 10:343±350

Weihs D, Webb PW (1984) Optimal avoidance and evasion tacticsin predator-prey interactions. J Theor Biol 106:189±206

Westby GWM, Keay KA, Redgrave P, Dean P, Bannister M (1990)Outpathways from the rat superior colliculus mediating ap-proach and avoidance have different sensory properties. ExpBrain Res 81:626±638

Yager D, Sharma SC, Grover BC (1977) Visual function in goldfishwith unilateral and bilateral tectal ablation. Brain Res 137:267±275

Yeomans JS (1990) Principles of brain stimulation. Oxford Univer-sity Press, New York