organization and significance of neurons that detect ...miller/lise_ref1.pdfhorizontal axis and...

Organization and Significance ofNeurons That Detect Change of Visual

Depth in the Hawk Moth Manduca sexta

MARTINA WICKLEIN AND NICHOLAS J. STRAUSFELD*

Arizona Research Laboratories Division of Neurobiology, University of Arizona,Tucson, Arizona 85721

ABSTRACTVisual stimuli representing looming or receding objects can be decomposed into four param-

eters: change in luminance; increase or decrease of area; increase or decrease of object perimeterlength; and motion of the object’s perimeter or edge. This paper describes intracellular recordingsfrom visual neurons in the optic lobes of Manduca sexta that are selectively activated by certainof these parameters. Two classes of wide-field neurons have been identified that respond selec-tively to looming and receding stimuli. Class 1 cells respond to parameters of the image otherthan motion stimuli. They discriminate an approaching or receding disc from an outwardly orinwardly rotating spiral, being activated only by the disc and not by the spiral. Class 2 neuronsrespond to moving edges. They respond both to movement of the spiral and to an approaching orreceding disc. These two classes are further subdivided into neurons that are excited by imageexpansion (looming) and are inhibited by image contraction (antilooming). Class 2 neurons alsorespond to horizontal and vertical movement of gratings over the retina. Stimulating class 1 and2 neurons with white discs against a dark background results in the same activation as stimu-lation with dark discs against a white background, demonstrating that changes in luminanceplay no role in the detection of looming or antilooming. The present results show that the twotypes of looming-sensitive neurons in M. sexta use different mechanisms to detect the approachor retreat of an object. It is proposed that cardinal parameters for this are change of perimeterlength detected by class 1 neurons and expansion or contraction visual flow fields detected byclass 2 neurons. These two classes also differ with respect to their polarity, the former comprisingcentripetal cells from the optic lobes to the midbrain, the latter comprising centrifugal neuronsfrom the midbrain to the optic lobes. The significance of these arrangements with respect tohovering flight is discussed. J. Comp. Neurol. 424:356–376, 2000. © 2000 Wiley-Liss, Inc.

Indexing terms: looming neurons; insect vision; neuroanatomy; electrophysiology; hovering flight

Perception of depth is a key feature of vision. It isused to detect and avoid objects; to pursue visual tar-gets; to maintain distance from targets; and to discrim-inate among objects in the third dimension. Animalsemploy many strategies to compute depth perception.These include stereopsis, in which images from eachretina are combined centrally to provide representationof depth by computing binocular retinal disparities (Ju-lesz, 1972; Livingston and Hubel, 1988). Vergence isanother binocular mechanism that uses the degree ofrotation of each eye, measured by eye muscle spindles,to compute absolute depth of a perceived target (Judge,1991). Vergence could, in principle, be used by arthro-pods with movable eyes on stalks, although it appearsthat distance perception by triangulation is the methodemployed (Zeil et al., 1986).

Insect eyes are integrated in the head cuticle and arevirtually immobile. In almost all insect species, the dis-tance between the two eyes is not sufficient for stereopsis

Grant sponsor: NIH National Center for Research Resources; Grantnumber: RR08688; Grant sponsor: A. v. Humboldt Foundation (FeodorLynen stipend); Grant sponsor: University of Arizona’s NSF IntegrativeGraduate Education and Research Traineeship Program; Grant sponsor:NSF Plant-Insect Interactions Group.

Martina Wicklein’s current address is: the Salk Institute, 10010 TorreyPines Rd., La Jolla, CA 92037-1099.

*Correspondence to: Nicholas J. Strausfeld, Arizona Research Laborato-ries Division of Neurobiology, 611 Gould-Simpson, University of Arizona,Tucson, AZ 85721. E-mail: [email protected]

Received 2 December 1999; Revised 25 April 2000; Accepted 27 April2000

THE JOURNAL OF COMPARATIVE NEUROLOGY 424:356–376 (2000)

© 2000 WILEY-LISS, INC.

or vergence other than in odonate larvae and in the prey-ing mantis, which use true stereopsis (Baldus, 1926; Ros-sel, 1986). A special case of binocular vision is used bymale calliphorid flies, which possess an area of high reti-nal acuity that provides information about the centeredimage of a pursued target. This is usually a female fly, theimmediate proximity of which is detected when its imagefalls on both retinas (Collett and Land, 1975).

Both vertebrates and arthropods also employ monocularmechanisms for depth perception. Not yet demonstratedin insects, but known from human psychophysics, is depthperception by occlusion where a depth map is created byobjects that are close to the moving eye appearing to passin front of more distant objects, blocking them and rela-tively larger areas of the background from view. Distanceperception by motion parallax has been identified in in-sects, particularly in locusts (Collett, 1978; Erikson, 1980)and in honey bees (Srinivasan et al., 1989). This type ofdepth perception relies on a comparison of velocities andmotion directions at different planes in the visual scene. Itshould be noted that depth perception by occlusion or bymotion parallax requires information about self-motionand about background.

These strategies contrast with yet another mechanismfor detecting change in depth: that of looming. This is theperception of an approaching object, the image of whichexpands (or contracts) across the retinal surface. Loomingis also an interesting case of depth perception because itinvolves neither background features nor self-motion of

the observer. The only requirement is that the object hasto contrast against a background. On its own, loomingcannot be used to define a three-dimensional (3D) map ofa scene. However, the perception of a looming stimulusinitiates escape behavior by locusts and flies and is usedby nectar feeding moths and hummingbirds for maintain-ing station in front of a target. It is thus not surprisingthat looming-sensitive neurons can be found in insectsand vertebrates. Examples are in sphingid moths (Wick-lein, unpublished data), grasshoppers (Rind, 1999; Gabi-anni et al., 1999), flies (Holmqvist and Srinivasan, 1991),pigeons (Sun and Frost, 1999), and primates (Tanaka andSaito, 1989; Graziano et al., 1994)).

This account addresses the question of what features ofthe looming stimulus are detected by uniquely identifiableneurons and how such neurons might collaborate to pro-vide information that would regulate the flight motor tomaintain distance from a target. The paper describes ex-periments performed on the tobacco hawk moth Manducasexta—a crepuscular and nocturnal sphingid. Sphingidsare rapid and highly maneuverable flyers with large com-pound eyes, which hover in front of flowers where theyperform stationary flight while ingesting nectar. Typi-cally, a moth will approach a flower at high velocity andthen abruptly decelerate to arrive at a distance from theflower that just permits insertion of the proboscis. To feedsuccessfully, the moth maintains its position even if theflower moves, irrespective of the direction of movement.Flowers selected as targets are those that offer stark con-trast in front of uniform foliage, which under these lightconditions comprises a virtually unstructured back-ground.

Fig. 1. Differences between stimuli used for this study. A: Theretinal image of an approaching (or receding) disc expands (or con-tracts) over the retina: the area of the image increases; the perimeterlength of the image increases; and the edge of the expanding imageprovides motion stimuli across ommatidia. B: A spiral that is fixed atone plane subtends an unchanging area of the retina. However, whenit turns, the edges of the spiral move across ommatidia, outward orinward depending on the direction of rotation. This stimulus providesmotion stimuli only, without changing perimeter length or imagearea.

Fig. 2. Histogram shows four examples of depth-sensitive neurons(mean responses of five repeated stimulations). Both class 1 and 2looming-sensitive neurons respond to an expanding disc with an in-crease of spike frequency (preferred direction) and a contracting (re-ceding) disc with a decrease of spontaneous spike activity (null direc-tion). The class 1 and 2 antilooming-sensitive neurons respond in theopposite manner: to a contracting disc with an increase of spikefrequency (preferred direction) and to an expanding (approaching)disc with a decrease of spontaneous spike activity (null direction). Thesymbols used here and in other figures are as follows: a wedge point-ing to the left indicates expansion; a wedge pointing to the rightindicates contraction.

357LOOMING-SENSITIVE NEURONS IN MANDUCA

MATERIALS AND METHODS

Animals

Manduca sexta (Lepidoptera: Sphingidae) were rearedon an artificial diet (modified after Bell and Joachim 1976)and maintained on a long-day photoperiod (17-hour light,7-hour dark) at 25°C at 50–60% relative humidity. Stage18 pupae were taken out of the colony and put into anenvironmental chamber to hatch. Their light cycle wasreversed in the chamber, and animals were kept underreversed conditions for at least a full cycle to ensure thatthey switched to a diurnal rhythm.

Electrophysiology: preparation

Adult 1- to 3-day-old M. sexta were cold anesthetizedand waxed to a holder to stabilize the thorax ventrally.The head was tilted forwards at a 30° angle from thehorizontal axis and waxed in place, leaving the eyes un-covered. The proboscis was cut to prevent muscle action,and the cut was sealed with low melting point wax. Anarrow window was cut into either the left or right poste-rior surface of the head capsule, directly above the opticpeduncle. Tracheal air sacs and the neural sheath werecarefully removed to expose the optic peduncle and themost medial part of the optic lobes. The preparation wasperfused with a Ringer’s solution to prevent desiccation(Christensen and Hildebrand, 1987). The head of the mothwas aligned in the rig in a standard fashion using theantennal bases and the base of the proboscis as fiducialpoints that were oriented to markers on the table. Theeyes’ pseudopupils were used to orient the head so that itassumed a standard position with respect to the stimuli,or the monitor, in front of it. To achieve this alignment thepseudopupils were revealed by illumination through anoperation microscope, and the head was moved until thepseudopupils were in the most medial and dorsal corner ofeach eye. These were then aligned with vertical and hor-izontal markers on the recording table.

Intracellular recording and staining

Intracellular recordings were made from neurons in theoptic lobes. Intracellular recording and dye injection wereperformed using silicon glass microelectrodes filled with a1% neurobiotin-KCl solution backed up with 1 M KCl. Theelectrodes had tip resistances of between 60 and 90 MV(when in contact with the tissue). A silver wire insertedinto the thorax served as the indifferent electrode. Record-ings were amplified (Axoprobe 1A, Axon Instruments,Burlingame, CA) and displayed using standard equipmentand stored (Video Recorder, Vetter, Rebersburg, PA). Af-ter recording, the cells were filled iontophoretically for5–10 minutes using a 1-nA negative current. Signals wereanalyzed off-line on a desktop computer using a data anal-ysis system (Spike2, CED, Cambridge, UK). The responsefrequency was calculated by subtracting the spontaneousactivity of the cell from the spike frequency during stim-ulation. An estimate of spontaneous activity was obtainedby calculating the means from the spike rate measured ina 2-second period before and after each set of motionstimuli.

Stimuli

Both physical objects (light or dark discs, gratings, ro-tating spirals) and their simulations were used as stimuli.All stimuli subtended a minimum of 30° on the retina.

Stimuli used to test looming responses were a white disc(5 cm in diameter) in front of a black background or ablack disc in front of a white background. The disc wasmoved with a constant velocity of 1 m/sec toward theanimal, stopped 10 cm in front of the moth where it sub-tended an angle of 30° on the retina, and then moved backto its starting position (30 cm away from the moth).

Spirals (10 cm in diameter) of different degrees of cur-vature, and presented at a standard distance from the eye,were rotated using a small motor that could be reversed(to provide apparent expansion or contraction). The spiralcould be rotated at a range of velocities. In each experi-ment, the spiral was rotated with constant velocity,stopped, and then rotated in the opposite direction withthe same velocity. Spiral stimuli were presented to thefront part of the visual field. To compare the effects ofdifferent stimuli, all the descriptions below refer to stim-ulation of this frontal field of view.

The stimulus used to test horizontal and vertical motionsensitivity consisted of a frame holding a movable grating.Gratings were moved vertically and horizontally in differ-ent parts of the visual field of the animal to test theresponse of a cell to translatory motion.

Stimuli were also simulated: displayed on a fast com-puter monitor (NEC M500, 160-Hz refresh rate) and pre-sented to the animal via a surface-glazed mirror posi-tioned to stimulate the frontal visual field (extending 70°in the horizontal plane and 50° in the vertical plane).These stimuli consisted of moving horizontal and verticalgratings; expanding and contracting light and dark discs;inwardly and outwardly rotating spirals. The starting andend size of the discs and the velocity with which theyexpanded and retreated could be preprogrammed, as couldthe size and curvature of the spiral and its velocity ofrotation.

Stationary stimuli were presented first, then moved inone direction, stopped, and moved in the opposite direc-tion. This cycle was repeated five times for each stimulus.The order in which stimulus types were presented wasrandomized.

Histology

The procedure for revealing neurobiotin was based onthe original method of Horikawa and Armstrong (1988),using a protocol for marine Crustacea (Schmidt and Ache1990) or modifications for Diptera (Gronenberg andStrausfeld, 1992). After iontophoresis (5–10 minutes 1nA)and 1 hour for tracer diffusion (at 4°C), the brain wasopened and immersed in 4% formaldehyde in 0.2 M So-rensen’s phosphate buffer. Half an hour later the brain

Fig. 3. Reponses of class 1 looming cells to expanding discs andspirals. Histograms (A,B) show relative spike frequencies elicited bya disc moving toward and away from the animal (wedges), a spiralturning outward and inward (indicated by arrowed spirals), horizon-tal progressive and regressive movements of gratings (horizontal ar-rows), and up-to-down or down-to-up motion of gratings (verticalarrows). The histograms show that both recorded neurons had thesame properties: strict tuning to disc expansion, inhibition to disccontraction, and no or negligible responses to directional motion.C–J: Spike trains supporting these conclusions. C,D: Excitation todisc expansion, inhibition to disc contraction. E,F: No responses toapparent looming (spiral). G–I: Movement of gratings shows no effecton background discharge.

358 M. WICKLEIN AND N.J. STRAUSFELD

Figure 3

was dissected and fixed overnight. After washing in bufferfor at least 8 hours, the tissue was dehydrated through anethanol series, rinsed in propylene oxide (10 minutes), andthen rehydrated. The purpose of this step is to permeabil-ize the tissue for a whole-mount avidin-biocytin procedure(Gronenberg and Strausfeld, 1992). The brain was thenincubated overnight in buffer with avidin-horseradish-peroxidase (Vectastain ABC-kit, Vector, Burlingame, CA)and 1% Triton X-100. After washing in buffer for 8 hours,the brain was presoaked for 4 hours in the dark at 4°C indiaminobenzidine (DAB) solution (Vectastain DAB-kit,Vector). Another buffer wash followed, and the tissue wasthen transferred into the same DAB solution containing0.001% hydrogen peroxide (Vectastain, DAB-kit, Vector)and processed in the dark at 4°C for 2 hours. After a1-hour rinse in buffer in the dark at room temperature,the brain was processed for an additional hour in the darkat room temperature. The tissue was rinsed in bufferovernight and then incubated in osmium tetroxide for2–12 hours in the dark at 4°C. The tissue was then dehy-drated and embedded in Durcupan (Fluka, Heidelberg,Germany).

Serial horizontal or frontal sections were cut at 35 mm.Preparations were observed and photographed with aLeitz Orthoplan microscope. Drawings and reconstruc-tions were made using a camera lucida attachment. Se-lected areas were serially digitized at 2-mm steps using aSony DC 5000 CCD and overlaid using Adobe Photoshop.Single images were collapsed into one image using thePhotoshop darkening function. Reconstructions have beennormalized to show one-half of the brain, as seen from theback or top, with the optic lobes to the left of the illustra-tion. Schematics showing the organization of the insectvisual system are provided in Strausfeld and Lee (1991).

RESULTS

More than 200 neurons were recorded from 148 moths.Most neurons were motion sensitive but not looming sen-sitive. Sixty-seven neurons that responded to loomingstimuli were recorded from the optic lobes. Of these, 14were filled with neurobiotin, with 14 others showing par-tial fills or multiple fills. The latter were discarded. Loom-ing neurons can be generally grouped into two majorclasses according to their tuning properties and structure.Class 1 neurons respond to approaching or receding discswhereas class 2 neurons respond both to these stimuli andto rotational motion of a spiral pattern (Fig. 1). Class 1and 2 neurons can be subdivided into those that are ex-cited by an object moving toward the eye (type 1a, 2aneurons) and those that are excited by an object recedingaway from the eye (type 1b, 2b neurons; Fig. 2A,B). Ofthose neurons that were filled with neurobiotin, class 1neurons were revealed to be centripetal cells leading fromthe optic lobes to the brain whereas class 2 neurons arecentrifugal cells, leading from centers in the posteriormidbrain (called the posterior slope) out to the medulla.The recorded neurons responded with action potentialshaving amplitudes that ranged from 10 to 60 mV. Re-sponses were mostly sustained tonic excitations or wereinhibition of background activity (Figs. 3A,B, 4A,B). Mostclass 2 neurons showed phasic-tonic responses and higherresponse frequencies at stimulus onset when stimulatedwith a rotating spiral (Figs. 4A,B, 9–11, 13).

Significantly, none of the neurons showed habituationto repetitive stimulation. Most neurons maintain a back-ground (resting) activity. Thus, suppression of back-ground in response to visual stimulation indicates inhibi-tion in the null direction. Conversely, excitation is anincrease of the spike frequency compared with the back-ground.

All the neurons tested had large (up to 120° horizontal,90° vertical) receptive fields in the frontal or frontoventralpart of the monocular visual field. Responses to a movingdisc or spiral were the same as those to movement simu-lated on a computer monitor.

Encoding of approaching and retreatingobjects

Neurons responded best to objects that move along thelongitudinal body axis of the moth toward or away fromthe eye (i.e., the z-axis). We classified neurons as loomingor antilooming sensitive and direction selective. Four ex-amples are shown in Figure 2. Two cells respond with anincrease in spike frequency when stimulated with a discmoving toward the eye and a decrease in spike frequencywhen a disc recedes from the eye (cells 1,2a). These neu-rons are classified as looming-sensitive neurons. Twoother neurons show an increase of spike frequency whenstimulated by a disc moving away from the eye and adecrease in spike frequency when the disc moves towardthe eye (cells 1, 2b). These cells exemplify antiloomingneurons.

Different types of stimuli that are associated withchange in distance reveal two classes of neurons thatrespond selectively to specific visual cues (Fig. 1): 1) ex-panding or contracting light and dark discs in front of acontrasting background provide change in object area,change of object perimeter (edge) length, and edge motion;2) inward and outward rotation of a spiral isolates edgemotion since this stimulus neither changes its area nor itsperimeter length; and 3) movement of horizontal and ver-tical gratings test for directionally selective motion sensi-tivity along the x and y axes.

Responses to these stimuli demonstrate that class 1 and2 neurons are sensitive to certain, but not all features ofan advancing or receding stimulus. Class 1 cells respondonly to an approaching or retreating disc but not to rotat-ing spirals or moving gratings. Class 2 cells respond to allthese different stimuli. Figures 3 and 4 exemplify re-sponses of two class 1 (Fig. 3) and two class 2 neurons (Fig.4) elicited by these stimulus elements. The bar graphs inFigure 3A and B summarize the response properties forthe two cell classes, each bar representing the mean re-sponse of five identical stimulations. The two class 1 neu-rons are clearly tuned to an expanding disc. Their spiketrains (Fig. 3C–H) show actual responses to differentstimuli: an approaching (expanding) disc causes excitation(Fig. 3D); a spiral simulating expansion elicits no response

Fig. 4. Reponses of class 2 looming cells to expanding discs andspirals. As in Figure 3, histograms A and boxes C, E, G, and I are fromone neuron; histogram B and boxes D, F, H, and J are from a secondneuron. Both are class 2 elements that are strongly excited by discexpansion, outward spiral rotation, regressive motion, and down-to-up motion. The cells are strongly inhibited by the respective oppo-site directions of motion or apparent motion.


Figure 4

(Fig. 3E,F); and likewise, horizontal (Fig. 3G,H) and ver-tical movement of gratings (Fig. 3I,J) are ineffective. Class1 neurons are further divided into two types: the 1a neu-rons just described are looming neurons, excited by anapproaching object and inhibited by a receding one; andthe type 1b neurons are antilooming neurons, inhibited byan expanding object and excited by a receding one. Bothtypes of neurons respond feebly, if at all, when presentedwith moving gratings.

Class 1 neurons distinguish expansion of the retinalimage from contraction. Contraction and expansion in-volve three variables: change of the perimeter length (edgelength) over the retina; change in the area of the image onthe retina; and change of luminance. Change in luminanceprobably plays no role in looming or antilooming detec-tion, however. A type 1a neuron that is activated by anexpansion of a white disc against a dark background isequally activated by an expansion of a dark disc against apale background (Fig. 5), where the overall luminancedecreases.

Two cells, the activities of which are shown in Figure 4,exemplify responses of class 2 neurons. The bar graphagain summarizes the response properties for each of thetwo cells; each bar represents the mean response of fiverepeated stimuli. The spike trains show responses of theneurons to the same stimuli used to test class 1 cells (Fig.3). Figure 4A and B demonstrates that class 2 cells aresensitive to expanding and receding discs. Here, both cellsare type 2a looming neurons, being excited by expansionand inhibited by contraction (Fig. 4B,C). The neuronsincrease their firing rates to outwardly rotating spirals(expansion) and decrease their firing rates to inwardlyrotating spirals (Fig. 4A,B,E,F). Thus, as in humans, theoutwardly rotating spiral provides the moth with the illu-sion of expansion. Again, class 2 neurons are divisible intotwo types: 2a responding to looming stimuli, and type 2bresponding to receding (antilooming) stimuli.

Class 2 neurons also respond to horizontal and verticalmovement of gratings over the retina (Fig. 4G–J). Thepreferred directions of the neuron summarized in Figure4A are front-to-back (regressive horizontal movement;Fig. 4G) and down-to-up vertical movement. The secondclass 2 neuron (Fig. 4B) preferred downward vertical mo-tion and regressive horizontal motion (Fig. 4I). In eachcase null directions were indicated by a reduction of theresting frequency.

Class 2 neurons were also tested with dark discs ex-panding or contracting against a pale background. As isthe case for class 1 neurons, the cells respond to changesof image size regardless of whether the stimulus is darkeror lighter than the background—in other words, indepen-dently of luminance (Fig. 5).

Anatomical differences of class 1and 2 neurons

The two classes of looming and antilooming neuronsdiffer in their morphology (Figs. 6–14). Class 1 neuronsderive from cell bodies in the optic lobes. Their terminalsinvade the ipsilateral optic foci of the posterior slope of themidbrain. Their dendrites in the optic lobes are groupedinto three distinct arbors: one in the medulla resides ex-clusively in its innermost stratum; a second in the lobulaconsists of processes that ascend through the lobula toprovide minute twigs within its outer stratum; and a thirdin the lobula plate penetrates all its strata.

Class 2 neurons derive from large cell bodies situatedposteriorly in the cell body rind between the calyces of themushroom bodies. The dendritic trees of class 2 neuronsreside in posterior slope neuropil on one or both sides ofthe brain. Their axons extend out to the medulla on thesame or the opposite side of the brain as their cell bodies.Their terminals in the optic lobes are restricted to theoutermost medulla stratum, extending across a large areaof its retinotopic mosaic.

Descriptions of class 1 neurons

Figure 6 shows a class 1 neuron filled with neurobiotinand reconstructed from serial sections in the verticalplane (normal to the z-axis). The cell responds with ahigher spike frequency to a receding stimulus but no ac-tivation when stimulated with an inward rotating spiral,thereby indicating that it is an antilooming neuron. Itsdendritic branchlets (Fig. 7A) are arranged distoproxi-mally in the innermost stratum of the medulla and in theoutermost stratum of the lobula where they arise fromstouter branches that ascend through this neuropil (Fig.7B). A third level of dendrites branch throughout thedepth of the lobula plate (Fig. 7C).

Figure 8A and B shows a second class 1 neuron, thedendrites of which are concentrated mainly within the lowerhalf of the medulla and lobula plate, suggesting that it isactivated by stimuli subtending the lower half of the visualhemisphere. However, its dendritic field in the lobula ex-tends across the dorsoventral extent of this neuropil. Theneuron terminates in the posterior slope of the midbrain.Original spike trains are shown in response to stimulationwith an approaching or retreating disc (Fig. 8B) and anoutward or inward rotating spiral (Fig. 8C). These are thetypical response characteristics of type 1b antilooming neu-rons, showing selective activation to a receding stimulus butno response to an inwardly rotating spiral.

Figure 9A illustrates a species of neuron that can be char-acterized as a type 1a looming cell, with a sustained responseto an expanding image on the retina (Fig. 9B). It is stronglyinhibited by a contracting image (Fig. 9B). An expandingspiral may slightly diminish background frequency, al-though this is ambiguous (Fig. 9C). Dendrites in the medulla

Fig. 5. Class 2 type 2 neuron, excited by disc contraction indepen-dent of luminance. Contraction of a white disc against a dark back-ground (white columns) excited the cell, as does expansion of a darkdisc against a light background (filled columns).


serve the lower frontal part of the eye’s receptive field as dodendrites in the lobula plate. Dendrites in the lobula areconcentrated in its ventral half and have diffuse arrange-ments dorsally. An axon projects to neuropil flanking the

esophagus. A thin collateral leads off the axon to provide arecurrent field of terminal processes situated at the base ofthe optic peduncle (Fig. 9A, inset). Its terminal in the poste-rior slope was incompletely filled.

Fig. 6. Reconstruction and responses of a class 1 type 2 antiloom-ing neuron. A: To show the full organization of this cell, the lobula isshown separated to the right from the medulla and lobula plate.Reconstruction of dendrites showing their distoproximal arrange-ments at three consecutive levels: in the medulla, lobula, and lobulaplate. Main branches ascend through the second chiasma to provideretinotopic branchlets in the inner medulla stratum and through the

lobula plate. Branches ascend outward through the depth of thelobula to its outer stratum, where they terminate. The inset shows anenlargement of medulla arborizations. Abbreviations for this andsubsequent figures: me, medulla; lo, lobula; lop, lobula plate. B: Re-cording showing the response of this cell to an expanding and con-tracting disc and (C) to an outward rotating spiral. Scale bar 5 100mm.


Descriptions of class 2 neurons

The next four figures illustrate class 2 neurons, all ofwhich have their cell bodies and dendrites in the mid-brain, as exemplified by the cell shown in Figure 10. Thiscell, which has been reconstructed from serial horizontalsections, has its terminal arborization confined to the out-ermost layer of the medulla. Its dendrites reside in thelateral deutocerebrum but originate from a large cell bodysituated contralaterally in the dorsal cell body rind. Thefine structure of the arborizations in the brain are sugges-tive of dendrites, indicating that this and other class 2neurons receive their inputs in the midbrain (see Fig. 12for details). This neuron showed spontaneous bursts ofactivity, which were possibly autonomous to this cell typebut are more plausibly a consequence of injury and cur-rent leakage. The neuron responded to an expanding ret-inal image and was inhibited by a contracting one. Itshowed an increase of spike frequency when stimulatedwith an outward rotating spiral (illusory looming). Theinitial response to this stimulus was phasic

The neurons shown in Figures 11, 13, and 14 werereconstructed from frontal sections. Figure 11 shows aneuron whose dendritic tree and cell body are contralat-eral to its terminal in the medulla. The latter provides adense field of beaded processes (Fig. 12A) that extendacross the whole outermost stratum of the medulla. Thedendrites branch profusely in the posterior slope of themidbrain and some extend distally into the base of theoptic peduncle. Dendritic branches are visited by axonsoriginating from the lobula plate (Figs. 11A, inset, 12B,C).The neuron is a looming cell showing an increase in activ-ity in response to an approaching disc or an outwardlyrotating spiral.

Figure 13 shows a class 2 cell that has two fields of den-drites, one on the same side as its terminal in the medulla,but contralateral to the cell body. Its axon prolongates overthe esophageal foramen to enter the posterior slope. Neuro-biotin failed to resolve details of the dendritic tree on thatside, however. As in the previous examples, a dense systemof terminal processes extends across the outer stratum of themedulla. The cell is a looming-sensitive neuron, typical ofother type 2a cells, responding to both an expanding disc andoutward spiral motion (Fig. 13A,B). The cell is completelysilent during stimulation with the nonpreferred direction ofmovement, namely, the retreating disc or the inwardly ro-tating spiral.

The last example (Fig. 14) shows a type 2a neuron withdendrites in the posterior slope of the midbrain. Its cellbody lies just at the other side of the midline, however.The terminal processes in the medulla are restricted to thelower half of the neuropil, in an area corresponding to thelower half of the retina. Figure 14B and C shows that theneuron is excited by an expanding retinal image and isinhibited by a contracting image. An outwardly rotatingspiral vigorously excites the cell.

DISCUSSION

This paper describes neurons in the visual neuropil ofManduca sexta that are tuned to looming stimuli. Two cellclasses can be distinguished, based on their anatomy andtheir visually induced response properties. Class 1 neu-rons receive inputs in the inner stratum of the medulla,the outer stratum of the lobula and through all strata of

the lobula plate. Their cell bodies are in the optic lobes,and their axons end ipsilaterally in the posterior slope ofthe midbrain. Class 2 cells are centrifugal neurons. Theyreceive their inputs in the posterior slope and terminate inthe medulla. Their cell bodies reside posteriorly near themidline of the brain. The two classes of neurons also differin their physiological responses to visual stimuli. Class 1cells respond only to an approaching or retreating disc butnot to a rotating spiral or moving grating. Class 2 cellsrespond to both types of stimuli. The different responseproperties of the neurons suggest that they detect differ-ent stimulus properties associated with looming.

In nature, tobacco hawk moths feed while they hover infront of a flower that is selected by the moth for its highcontrast (Raguso et al., unpublished data). While feeding,moths maintain a constant distance from the flower usingvisual cues (Wicklein and Willis, unpublished data). Threefeatures of the target could provide information that isused by the visual system to detect change of distance: 1)change in the area of the retinal image; 2) change inlength of the image perimeter; and 3) detection of flowfield properties, such as expanding or contracting edges.The present results demonstrate that at least two of thesemechanisms are realized within one and the same system.

Two components of the stimulus can provide informa-tion about change in area: a change in luminance detectedby pooling the intensity signal across the whole retina; ordetection of change of image size on the retina by recruit-ment or divestment of individually stimulated retinotopicchannels. Cells that increase their firing rate to an ex-panding bright target against a darker background alsorespond in the same way to an expanding dark targetagainst a bright background. This excludes the possibilitythat the effective signal is increased luminance becausethe second stimulus condition is associated with an overallluminance decrease.

Do changes in perimeter length of the image provideinformation about looming? To test this, neurons werestimulated with expanding and contracting discs. Theirresponses were compared with responses elicited by in-wardly and outwardly rotating spirals. A looming objectexpanding over the retina’s surface not only provides out-ward moving edges but also increasing area and an in-crease of perimeter (edge) length. A spiral rotated in onedirection provides an outward motion of edges and in theopposite direction an inward motion of edges, in both caseswithout any change in area or perimeter length. Thus, ifperimeter length is a key stimulus for looming detection,an outwardly rotating spiral should not excite a cell that isexcited by an expanding disc.

Fig. 7. Features of the neuron shown in Figure 6. A: Tufts arerestricted to the innermost stratum (is) of the medulla’s inner layer(il). The outer layer (ol) and serpentine layer (sl) are not invaded bythis neuron. B: Detail showing branches ascending through the innermedulla strata (is). C: Branches also arborize through the outer (ol)and inner (il) layers of the lobula. The inset (from the boxed area)shows delicate brushlets of processes that invade the outer lobulalayer, at the level of terminals of type 1 transmedullary cells from themedulla (Fig. 15). D: A third system of branches invades the lobulaplate, extending from its outer surface through all three direction-and orientation-specific layers (1–3). p op t; groove in the outer surfaceof the lobula carrying axons of the posterior optic tract. Scale bars 550 mm in A; 25 mm in B; 10 mm in C,D.


Figure 7

The present results show that only class 1 neurons areactivated by expanding or contracting discs but not bythe rotating spiral. We conclude that these neurons aretuned to changes of retinal image area or changes ofperimeter (edge) length and that luminance and movingedges play no role in their activation. Class 2 neuronsrespond to both expanding and contracting discs and torotating spirals. This shows that outward and inwardmovement of edges are their effective stimuli and thatluminance, perimeter length, and changes of retinalimage area are not necessary to excite this class of cells.In addition to their responses to spirals, class 2 cells areactivated by moving gratings whereas class 1 cells arenot. This supports the idea that class 2 cells are tunedto directional motion whereas class 1 cells are not. Re-sponses of class 2 cells to changes of image size are

independent of luminance. This fits well with naturalconditions. A white flower illuminated by moonlight orstarlight will appear pale against a dark background.The same flower may offer a dark profile against a palebackground (the sky) at and just after sunset when themoths are already foraging.

Anatomical polarities of class 1and 2 neurons

Both class 1 and 2 neurons are conventional nervecells in that they have two separate fields of arboriza-tions connected by an axon. Arborizations that are in-terpreted as dendrites can be either smooth or tapering,or can comprise many tapering branchlets equippedwith knobs, spines, or varicosities. Structures inter-

Fig. 8. A class 1 type 2 antilooming neuron. A: Localized dendriticfields in the ventral medulla and lobula plate and across the lobula.The axons terminate in the posterior slope (p sl). Other abbreviations:la, lamina; mb, calyx and initial part of pedunculus of the mushroom

body; p sl, posterior slope. B: Responses to disc expansion (inhibition)and contraction (excitation) and (C) an unchanged background activ-ity during stimulation with an expanding (or contracting: not shown)spiral. Scale bars 5 100 mm.


preted as terminals are composed of arborizations dec-orated with varicosities or bead-like specializations(Strausfeld and Meinertzhagen, 1998). Reconstructionsof class 1 and 2 neurons suggest that the two classes aredifferently polarized: class 1 neurons are centripetal,carrying information centrally from dendrites in opticlobe neuropils to terminals in the midbrain; class 2

neurons are centrifugal, relaying information from theirdendrites in the posterior slope out to their terminals inthe medulla. The locations of their cell bodies supportthese polarities: class 1 neurons have cell bodies in theoptic lobes, and class 2 neurons have their cell bodies inthe posterior rind of the midbrain. One caveat to theseanatomical inferences is that neurons in insects, as in

Fig. 9. A class 1 type 1 looming neuron. A: Reconstruction showingthat its processes are arranged across the central part of the medulla’sretinotopic mosaic. Processes are distributed across the whole lobulaand lobula plate. The terminals arborize extensively in the neuropil,connecting the brain with the subesophageal ganglion. A small col-lateral from the axon provides a field of blebbed processes (boxed; see

inset) within the optic peduncle, suggesting their relationship withaxons of other visual interneurons. B: Spike train showing excitationelicited by an expanding disc and inhibition to a contracting disc.C: There are no responses to an expanding spiral. Scale bar 5 100 mmin A; 5 mm in inset.


deuterostomes, can have pre- and postsynaptic struc-tures on the same arbor. We were not able to performphysiological experiments confirming the assumed di-rection of information flow.

Centrifugal cells have been previously described fromthe butterfly Papilio aegeus (Ibbotson et al., 1991) andfrom Manduca sexta (Milde, 1993). Ibbotson et al. (1991)found motion-sensitive neurons that connect the midbrainto the medulla. These neurons (termed MV1 and MH1)have cell bodies close to the midline of the brain. Theirsmooth tapering arborizations are located in the posteriorslope, and their wide-field terminals reside in the outerlayers of the medulla. These cells correspond structurallyto the class 2 neurons described here. In Papilio, MH1cells are selectively activated by horizontal motion with apreferred and null direction. MV1 neurons respond tovertical motion, also with a preferred and null direction.Interestingly, the MH1 and MV1 cells of Papilio are tunedonly to horizontal or to vertical gratings, respectively,whereas the present class 2 cells in Manduca are sensitiveto movement in both orientations, with a null and pre-ferred direction for each. Cells termed cMt by Milde (1993)have their cell bodies located dorsally, close to the brain’s

midline. Like the class 2 neurons described here, cMtneurons have smooth tapering processes in the posteriorslope and beaded arborizations in the outer third of themedulla. Whereas processes of class 2 cells extendthrough large areas of the medulla’s retinotopic mosaic,the terminals of cMt cells are restricted to small patches ofthe medulla. The cMt neurons are motion sensitive andsome show directional selectivity (Milde, 1993). NeithercMt nor MH1 and MV1 neurons were tested with loomingstimuli and thus cannot be further compared with class 2neurons described here. We assume that all theseneurons—class 2 cells, cMt, MH1, and MV1 neurons—belong to a larger category of motion-sensitive elementsthat are supplied by afferents in the posterior slope andfeed information back to the medulla. The nature of theirinputs is discussed next.

Polarities of looming neurons

Observations of dipteran visual systems reveal that theposterior slope of the midbrain is the main terminationarea for wide-field motion-sensitive collator neurons orig-inating in the lobula plate (Strausfeld and Bassemir,1985a,b; Strausfeld and Gronenberg, 1990). These direc-

Fig. 10. Class 2 type 1 neuron. A: Reconstruction from serialhorizontal sections shows the peripheral location of its terminals atthe medulla’s outer surface and a densely branched dendritic treeipsilaterally in the midbrain and its contralateral cell body. ant lob,

antennal lobe. Other abbreviations as for previous figures. Althoughbackground bursts of spikes may be injury responses, the cell never-theless showed a clear response to disc expansion (B) and to spiralexpansion (C). Scale bar 5 100 mm.


tionally selective neurons, which are variously tuned tospecific orientations of motion, provide inputs to the den-dritic trees of premotor descending neurons and arethought to be also presynaptic to a variety of centrifugaland heterolateral cells that extend back out into the opticlobes (Hausen and Egelhaaf, 1989; Strausfeld et al., 1995).The present results suggest that class 2 neurons are com-parable to centrifugal neurons in Diptera, like the CHneuron (Hausen and Egelhaaf, 1989). Moreover, observa-tions of biocytin-filled class 2 neurons reveal that theirdendrites in the posterior slope are visited by many large-diameter profiles from the lobula plate.

Given this anatomical relationship, it is not surprisingthat class 2 neurons are orientation and directionally mo-tion sensitive. They respond to a preferred vertical as wellas to a preferred horizontal direction, suggesting thattheir dendrites receive inputs from two converging popu-

lations of lobula plate outputs: horizontal motion-sensitiveneurons and vertical motion-sensitive neurons. Recentstudies of the diurnal hummingbird hawk moth Macro-glossum stellatarum (Wicklein and Varju, 1999) have con-firmed that the sphingid lobula plate indeed providesseparate populations of vertical and horizontal motion-sensitive tangential neurons that are probably homolo-gous to the horizontal and vertical motion-sensitive neu-rons of many brachyceran Diptera (Buschbeck andStrausfeld, 1997).

Organization among the dendrites of class 1 neurons ismore difficult to interpret. The entire dendritic tree pro-vides three orderly and spatially segregated systems ofbranches that supply distal processes to three discretelevels of optic lobe: the inner stratum of the medulla; theoutermost stratum of the lobula; and all strata throughthe lobula plate. Each branching system originates from a

Fig. 11. A class 2 type 1 neuron reconstructed from vertical sec-tions. A: Its large dendritic tree in the right side of the brain providesan axon that terminates in the lower frontal two-thirds of the con-tralateral medulla, there providing a dense plexus of beaded pro-cesses. Its dendritic tree is subdivided into branchlets (inset, upperright) each providing a claw-like group of processes that are investedwith knobs and spicules. These appear to clasp the passage of un-

stained axons that project from the right lobula complex (not shown)to the posterior slope. B: This neuron showed a high level of autono-mous activity. The firing rate increased in response to an expandingdisc but returned to its initial rate in response to a contracting disc.C: Its firing rate increased in response to an outwardly rotating spiral.Scale bar 5 100 mm.


Fig. 12. Details of the class 2 neuron shown in Figure 11.A: Terminal processes are equipped with bead-like specializations(arrowed circle) that invade retinotopic columns. B: This panel showsan enlargement of the area boxed to the right in Figure 11, identifying

dendritic processes in the posterior slope. C: Enlargements showingthe clawed appearance of dendritic branchlets (arrows and boxes).D: Enlargement of dendritic tuft boxed to the right in B. Scale bars 510 mm.


large diameter axon. Branches are ordered: those supplyingthe medulla are distal to branches supplying the lobula;these are distal to branches supplying the lobula plate.

In sphingids, as in brachyceran Diptera, levels receivingclass 1 neuron dendrites are specifically associated withsmall-field retinotopic neurons that contribute to a dis-crete motion processing pathway (Douglass and Straus-feld, 1995, 1997) that is composed of elementarymovement-detecting circuits (EMDs; Buchner, 1976).These circuits are repeated across the optic lobes, eachcircuit detecting the sequential change of luminance be-tween adjacent visual sampling points (Franceschini etal., 1989). An integral part of each EMD is the type 1transmedullary neuron (Tm1), which occurs in each reti-notopic column. Tm1 neurons, which are common to bothflies and sphingid moths (Fig. 15), are modulated by di-rectional motion (Douglass and Strausfeld, 1995). Their

collaterals invade the innermost medullary stratum, andtheir endings terminate in the outermost stratum of thelobula (Fig. 15). Tm1 neurons supply information aboutmovement to the next processing stage, the bushy T-cells(T5), whose dendrites reside in the outer stratum of thelobula, as do dendritic twigs of class 1 neurons. BushyT-cells terminate in orientation- and direction-specificstrata in the lobula plate. Again, class 1 neuron dendritesinvade these layers.

The coincidence of three levels of class 1 dendrites andthe respective levels of Tm1 collaterals, Tm1 terminals,and T5 endings suggests that class 1 neuron dendritesreceive inputs from movement-sensitive neurons acrossthe whole eye. Why, though, should this occur thrice, atsuccessive levels in the system? And why should class 1neurons be “blind” to translatory motion if they are sup-plied by motion-sensitive elements?

Fig. 13. Reconstruction of a class 2 type 1 neuron showing its terminal in the outer stratum of themedulla. A: The bilateral dendritic tree is shown only in part with the right component indicated by thedotted line. B: The neuron showed almost no background activity and was excited by disc expansion andoutward spiral motion (C). Scale bar 5 100 mm.


One suggestion is that the relationship between Tm1 neu-rons and class 1 neuron dendrites in the medulla and outerlobula allows successive temporal sampling of changes ofcontrast between adjacent ommatidia, as would occur whenthe image perimeter expands or contracts across the retina.In the medulla and lobula, each dendritic branchlet of a class1 neuron thus serves as an edge detector. The neurons’ lackof response to direction may be explained by directionalinformation being discarded due to interactions in the lobulaplate where class 1 neuron dendrites invade all its direction-ally sensitive layers.

The question of why dendrites of class 1 neurons aredistributed successively within the three optic lobe neuro-pils is more challenging. One possibility is that duplicat-ing inputs from Tm1 at two levels adds redundancy to thesystem and thus increases accuracy. Alternatively, den-drites staggered at three successive levels may enable thecell to detect temporal and spatial displacements of theimage edge from one ommatidium to the next during im-

age expansion or contraction. Estimations of change oflength of the image edge as it contracts or expands acrossthe retina will be a function of the number of edge detec-tors triggered by edge motion. This calculation can ill

Fig. 15. Camera lucida drawings of small-field retinotopic neuronshomologous to those that are known to be involved in elementarymotion-detecting circuits in Diptera and that terminate in layers thatare invaded by the dendrites of class 1 looming neurons. Small andwide-field lamina monopolar cells (L2s, w) terminate in the outerstratum of the medulla, at the level of terminals of class 2 loomingneurons and the dendrites of type 1 small and wide-field transmed-ullary neurons (Tm1 s,w). These second-order neurons provide narrowor wide-field collaterals within the inner stratum of the medulla andterminals in the outer stratum of the lobula. The dendrites of bushyT-cells (T4; see inset for details, T5) originate at these levels, eachsending an axon into motion- and orientation-specific levels in thelobula plate (lop). These layers, like the inner stratum of the medullaand outer stratum of the lobula are invaded by the class 1 neurondendrites. Scale bar 5 100 mm.

Fig. 14. A class 2 type 1 neuron with a dendritic tree in the ipsilateral posterior slope, but derivedfrom a contralateral cell body. A: The terminal in the medulla invades only the lower half of theretinotopic mosaic. B: This neuron responded to an expanding disc and (C) to an outwardly expandingspiral. Scale bar 5 100 mm.


Figure 15

afford noise incurred by synaptic delays, as velocities ofimage edge expansion/contraction over the retina whilehovering are fast, calculated as 0.2–0.4 m/sec. Possibly,activated edge detectors are pooled first in the medulla,then in the lobula and then in the lobula plate. The orga-nization of class 1 dendrites at these levels could allowcomparison of rapid change in perimeter distribution onthe retina by integrating edge detection within the den-dritic tree at time t1 for inputs to the medulla, t2 for inputsto the lobula, and t3 for inputs to the lobula plate.

Why are there looming and antiloomingsystems? Comparisons with other taxa

The present results identify two classes of neurons thatdetect movement along the z-axis: cells that are insensi-tive (class 1) or sensitive to motion cues (class 2). Bothclasses are further divided into two types (1, looming; 2,antilooming) that are excited by expansion or contraction,respectively, of a retinal image. If each of these cell typescan alone signal change of depth, what is the advantage ofhaving a system comprising four types of neurons thatsignal the same event?

In locusts and flies a single system of neurons detectslooming to provide information about the expansion of animage over the retina. This information triggers an escapejump or a landing response, depending on the context inwhich the stimulus is encountered. In locusts this behav-ior is mediated, in part, by the lobula giant motion detec-tor (LGMD), which synapses onto the descending con-tralateral motion detector (DCMD: Rowell, 1971; Rind,1984)—itself supplying the hind-leg jump circuit. TheLGMD responds to a looming stimulus that may (Gabbi-ani et al., 1999) or may not (Rind and Simmons, 1999)signal time-to-collision.

In flies there is an isomorphic array of several hundredcolumnar neurons called Col A cells that subserve thewhole eye. Each cell is triggered by a darkening stimulusentering its receptive field (Gilbert and Strausfeld, 1991),and the axon of each Col A cell is coupled by gap junctionsto dendrites of the giant descending neuron (GDN) thatsupplies the midleg jump circuit (Strausfeld and Bas-semir, 1983). This system reacts to a stimulus thresholddefined by perimeter length of the retinal image ratherthan to time-to-collision (Holmqvist and Srinivasan,1992).

Swimming, flying, or arboreal animals would be ex-pected to possess neuronal circuits that inform motorpathways about impending collision with objects. Suchcircuits are not exclusive to insects. They have been pro-posed for crustaceans (Glantz, 1974), and these circuitsare likely to have evolved in arthropods before their ap-pearance in chordates. Neurons having similar morphol-ogies to those seen in insect looming systems have beenidentified in the nucleus rotundus of the pigeon (Luksch etal., 1998) and in layer SGS1 of the squirrel optic tectum(Major, Luksch and Karten, unpublished data). In eachcase the dendrites of these cells provide broad fan-likearrays of dendritic branches, each providing a smallbottlebrush-like tuft onto which centripetal afferents ter-minate. As in the class 1 looming neurons, the shapes ofthese elements suggest a precise retinotopic input ontothem from quite large areas of the visual field, which is

also the case for the locust LGMD and the ensemble of ColA cells in flies.

There are interesting parallels between the selectiveproperties of wide-field neurons in the looming subdivi-sions of the nucleus rotundus of pigeons (Wang andFrost, 1992; Wang et al., 1993) and type 1 class 1 andtype 1 class 2 neurons in Manduca. In the nucleusrotundus, wide-field neurons are segregated at two dis-crete levels, the more distal cells receptive to velocity onthe x and y axes, the deeper neurons tuned to motion onthe z axis (Wang et al., 1993). Like class 1 cells, theseneurons are not sensitive to directional motion but dorespond to image expansion. Sun and Frost (1998) showthree types of looming-sensitive neurons in the nucleusrotundus: one computing the relative rate of expansion;the second the absolute rate of expansion; and a thirdmultiplying size-dependent responses and instanta-neous angular velocity of the object in a manner com-parable to findings from the locust LGMD (compareGabbiani et al., 1999).

What is so remarkable about these various systems isthe apparent evolutionary convergence—both of morpho-logical attributes and of functional properties—suggestingthat common selective pressures, such as the necessity ofavoiding objects during fast movements in three dimen-sions (water or air), have forced similar adaptations acrossphyla. It would be most interesting to study the nucleusrotundus of hummingbirds because, unlike pigeons, hum-mingbirds use looming and receding stimuli as cues forstationary flight. In this they show an exquisite parallel tosphingid hovering, a behavior that is utterly distinct fromthe escape and landing responses of locusts that relysolely on looming stimuli.

There are similarities between the organization ofcontrol systems for stationary and for forward flight.During stationary flight the flight motor continuouslymakes compensatory fine adjustments to adjust for vi-sual drift—signifying changing position toward or awayfrom the target (Pfaff and Varju, 1991; Farina et al.,1994, 1995; Kern and Varju, 1998). In a comparablemanner, compensatory movements stabilize the plat-form along the pitch, yaw, and roll axes to maintainvisual balance during forward flight (Hausen andEgelhaaf, 1989). Although it could be argued that forforward flight one system would suffice to detect pro-gressive and regressive horizontal motion, it seems thatthis is insufficient for fine control: several orthogonalsystems have evolved that provide highly accurate in-formation about all extractable features of the visualflow field (Hausen, 1982; Krapp and Hengstenberg,1996). The several systems supporting stationary flightare likewise organized as discrete assemblages of nervecells that collaborate in the detection of movementalong the z axis. Even though systems of looming neu-rons in moths are crudely distinguished as motion-independent and motion-dependent neurons, their fourdivisions into looming and antilooming neurons de-scribed here may not be the minimum required forprecise control of stationary flight. In this account wehave shown that for any type of looming or antiloomingneuron there exist several distinct morphs or species ofnerve cells that have different field dimensions acrossthe neuropil. Possibly, many of these morphologicaltypes are required to provide the exquisite accuracy ofmotor control manifested by the moth’s behavior. Com-


parative studies of insects that show different grades ofvisually guided behaviors may resolve the question ofwhether systems of looming and antilooming neuronsderive from an evolutionary elaboration of a simplerand phylogenetically basal system, like the locustLGMD.

ACKNOWLEDGMENTS

We thank Albert Brower, BS, for instructions in imag-ing and digital montaging, Robert Gomez, BS, for techni-cal help, and A.A. Osman, PhD, for rearing moths. We aregrateful to Holly Campbell, PhD, for discussing variousaspects of this work. This account has also profited fromHarvey Karten’s (MD) valuable insights about avian vi-sual systems. N.J.S. received a grant from the NIH Na-tional Center for Research Resources (RR08688), andM.W. received a Feodor Lynen stipend from the A. v.Humboldt Foundation, as well as small grants from theUniversity of Arizona’s NSF Integrative Graduate Educa-tion and Research Traineeship Program and NSF Plant-Insect Interactions Group.

LITERATURE CITED

Baldus K. 1926. Experimentelle Untersuchungen uber die Entfer-nungslokalisation der Libellen (Aeschna cyanae) Z Vergl Physiol 3:475–505.

Bell RA, Joachim FA. 1976. Techniques for rearing laboratory colonies forrearing tobacco hornworms and pink bollworms. Ann Ent Soc Am69:365–373.

Buchner E. 1976. Elementary movement detectors in an insect visualsystem. Biol Cybern 24:85–101.

Buschbeck EK, Strausfeld NJ. 1997. The relevance of neural architectureto visual performance: phylogenetic conservation and variation indipteran visual systems. J Comp Neurol 383:282–304.

Christensen TA, Hildebrand JG. 1987. Male-specific, sex pheromone-sensitive-selective projection neurons in the antennal lobes of the mothManduca sexta. J Comp Physiol A 160:553–569.

Collett TS. 1978. Peering—a locust behaviour pattern for obtaining motionparallax information. J Exp Biol 76:237–241.

Collett TS, Land MF. 1975. Visual control of flight behaviour in the hover-fly, Syritta pipiens L. J Comp Physiol (A) 99:1–66.

Douglass JK, Strausfeld NJ. 1995. Visual motion detection circuits in flies:peripheral motion computation by identified small field retinotopicneurons. J Neurosci 15:5596–5611.

Douglass JK, Strausfeld NJ. 1995. Visual motion detection circuits in flies:parallel direction- and non-direction-sensitive pathways between themedulla and lobula plate J Neurosci 16:4551–4562.

Eriksson ES. 1980. Movement parallax and distance perception in thegrasshopper (Phanlacridium vittatum) (Sjostedt) J Exp Biol 86:337–341.

Farina WM, Varju D, Zhou Y. 1994. The regulation of distance to dummyflowers during hovering flight in the hawk moth Macroglossum stella-tarum. J Comp Physiol 174:239–247.

Farina WM, Kramer D, Varju D. 1995. The response of the hovering hawkmoth Macroglossum stellatarum to translatory pattern motion. J CompPhysiol 176:551–562.

Franceschini N, Riehle A, Le Nestour A. 1989. Directionally selectivemotion detection by insect neurons. In: Stavenga DG, Hardie RC,editors. Facets of vision. Berlin: Springer. p 360–390.

Gabbiani F, Krapp H, Laurent G. 1999. Computation of object approach bya wide-field motion-sensitive neuron. J Neurosci 19:1122–1144.

Gilbert C, Strausfeld NJ. 1991. The functional organization of male-specificvisual neurons in flies. J Comp Physiol A 169:395–411.

Glantz, RM. 1974. Defense reflex and motion detector responsiveness toapproaching targets: the motion detector trigger to the defense reflexpathway. J Comp Physiol (A) 95:297–314.

Graziano, MSA, Andersen RA, Snowden RJ. 1994. Tuning of MST neuronsto spiral motions. J Neurosci 14:54–67.

Gronenberg W, Strausfeld NJ. 1992. Premotor descending neurons re-sponding selectively to local visual stimuli in flies. J Comp Neurol316:87–103.

Hausen K. 1982. Motion sensitive interneurons in the optomotor system ofthe fly. II. The horizontal cells: receptive field organization and re-sponse characteristics. Biol Cybern 46:67–79.

Hausen K, Egelhaaf M. 1989. Neural mechanisms of visual course controlin insects. In: Stavenga DG, Hardie R, editors. Facets of vision. NewYork: Springer. p 391–424.

Holmqvist MH, Srinivasan MV. 1991. A visually evoked escape response ofthe housefly. J Comp Physiol (A) 169:451–459.

Horikawa K, Armstrong WE. 1988. A versatile means of intracellularlabeling: injection of biocytin and its detection with avidin conjugates.J Neurosci Methods 25:1–11.

Ibbotson MR, Maddess T, Du Bois R. 1991. A system of insect neuronssensitive to horizontal and vertical image motion connects the medullawith the midbrain. J Comp Physiol (A) 169:355–367.

Judge SJ. 1991. Vergence. In: Carpenter RHS, editor. Eye movements, vol8. Vision and visual dysfunction (Cronly-Dillon JR, series editor). BocaRaton, FL; CRC Press. p 157–174.

Julesz B. 1972. Cyclopean perception and neurophysiology. Invest Oph-thalmol 11:540–548.

Kern R, Varju D. 1998. Visual position stabilization in the hummingbirdhawk moth, Macroglossum stellatarum L. I. Behavioral analysis.J Comp Physiol 128:225–237.

Lee DN, Davies MNO, Green PR, Van der Weel FR. 1993. Visual control ofvelocity of approach by pigeons when landing. J Exp Biol 180: 85–104.

Krapp H, Hengstenberg R. 1996. Estimation of self-motion by optic flowprocessing in single visual interneurons. Nature 384:463–466.

Livingstone M, Hubel D. 1988. Segregation of form, color, movement,and depth: anatomy, physiology, and perception. Science 240:740 –749.

Luksch H, Cox K, Karten HJ. 1998. Bottlebrush dendritic endings andlarge dendritic fields: motion-detecting neurons in the tectofugal path-way. J Comp Neurol 396:399–414.

Milde JJ. 1993. Tangential medulla neurons in the moth Manduca sexta.Structure and responses to optomotor stimuli. J Comp Physiol 173:783–799.

Pfaff M, Varju D. 1991. Mechanisms of visual distance perception inthe hawk moth Macroglossum stellatarum. Zool Jb Physiol 95:315–321.

Rind FC. 1984. A chemical synapse between two motion detecting neuronesin the locust brain. J Exp Biol 110:143–167.

Rind FC, Simmons PJ . 1999. Seeing what is coming: building collision-sensitive neurones. Trends Neurosci 22:215–220.

Rossel S. 1986. Binocular spatial localization in the praying mantis. J ExpBiol 120:265–281.

Rowell CHF. 1971. The orthopteran descending movement detector (DMD)neurones: a characterisation and review. Z Vgl Physiol 73:167–194.

Schmidt M, Ache BW. 1990. Afferent projections to the midbrain of thespiny revealed by biocytin. Soc Neurosci Abstr 16:400.

Srinivasan MV, Lehrer M, Zhang SW, Horridge GA. 1989. Honeybeesmeasure their distance from objects of unknown size. J Comp Physiol(A) 165:605–614.

Strausfeld NJ, Bassemir UK. 1983. Cobalt-coupled neurons of a giant fibresystem in Diptera. J Neurocytol 12:971–991.

Strausfeld NJ, Bassemir UK. 1985a. Lobula plate and ocellar interneu-rons converge onto a cluster of descending neurons leading to neckand leg neuropil in Calliphora erythrocephala. Cell Tissue Res 240:617– 640.

Strausfeld NJ, Bassemir UK. 1985b. The organisation of giant horizontal-motion-sensitive neurons and their synaptic relationships in the lateraldeutocerebrum of Calliphora erythrocephala and Musca domestica.Cell Tissue Res 242:531–550.

Strausfeld NJ, Gronenberg W. 1990. Descending neurons supplyingthe neck and flight motor of Diptera: organization and neuroana-tomical relationships with visual pathways. J Comp Neurol 302:954 –972.

Strausfeld NJ, Lee J-K. 1991. Neural basis for parallel visual processing inthe fly. Vis Neurosci 7:13–33.

Strausfeld NJ, Meinertzhagen IA. 1998. The insect neuron: types, mor-phologies, fine structure, and relationship to the architectonics of the


insect nervous system. In: Harrison FW, Locke M, editors. Microscopicanatomy of invertebrates. Insects, vol 11B. New York: Wiley-Liss. p487–538.

Strausfeld NJ, Kong A, Milde JJ, Gilbert C, Ramaiah L. 1995. Oculomotorcontrol in calliphorid flies: GABAergic organization in heterolateralinhibitory pathways. J Comp Neurol 361:298–320.

Sun H, Frost BJ. 1998. Computation of different optical variables of loom-ing objects in pigeon nucleus rotundus neurons. Nature Neurosci1:296–303.

Tanaka K, Saito H. 1989. Analysis of motion of the visual field by direction,expansion/contraction, and rotation cells clustered in the dorsal part of

the medial superior temporal area of the macaque monkey. J Neuro-physiol 62:626–641.

Wang YC, Frost BJ. 1992. Time to collision is signaled by neurons in thenucleus rotundus of pigeons. Nature 19:236–238.

Wang YC , Jiang S, Frost BJ. 1993. Visual processing in pigeon nucleusrotundus: luminance, color, motion, and looming subdivisions. Vis Neu-rosci 10:21–30.

Wicklein M, Varju D. 1999. The visual system of the European hummingbirdhawkmoth Macroglossum stellatarum (Sphingidae, Lepidoptera). Motionsensitive Interneurons of the lobula plate. J Comp Neurol 408:272–282.

Zeil J, Nalbach G, Nalbach H-O. 1986. Eyes, eye stalks and the visualworld of semi-terrestrial crabs. J Comp Physiol (A) 159:801–811.


organization and significance of neurons that detect ...miller/lise_ref1.pdfhorizontal axis and...

Documents