Avian Visual Cognition

Husband & Shimizu Home Page

Next Section:

References and Links



V. Taking Flight: Post-Retinal Processing 
Collo- & Lemno-Thalamic Visual Pathways 

Amniote visual pathways. The two major visual pathways found in reptiles, birds, and mammals show a high degree of anatomical similarity.  These two pathways arise from retinal efferents which travel to either the optic tectum (TeO in birds, superior colliculus in primates) or dorsal thalamus (OPT in birds, LGNd in primates).  The retino-collicular pathway has been called a tectofugal (Karten & Revzin, 1966) or collothalamic pathway (Butler & Hodos, 1996).  From the collicular structure there are efferents to the thalamus (nucleus Rotundus, Rt, in birds, LP/pulvinar in primates).  The thalamic nuclei send projections to the telencephalon (ectostriatum in birds, extrastriate cortex in primates).  Other retinal efferents travel to dorsal thalamic nuclei travel to areas in the telencephalon (visual wulst in birds, striate cortex in primates).  This retino-thalamo-striate pathway has been called a thalamofugal (Karten & Nauta, 1968) or lemnothalamic pathway (Butler & Hodos, 1996). 

While there are many similarities in the avian and mammalian visual pathways, the relative importance of each pathway to visual perception appears to be different.  The past three decades of work on mammalian visual systems has concentrated on the geniculo-striate system.  Studies on the lateral geniculate nucleus of the dorsal thalamus, stiate cortex (V1), and extrastriate areas (e.g., V4, MT) have provided a wealth of detail on the visual systems of mammals and how brains perform visual processing in general.  In birds, however, the situation is somewhat reversed.  The avian collothalamic system may be the dominant visual pathway, serving many of the functions ascribed to the primate lemnothalamic system.  This generalization appears to be true in at least lateral-eyed species like the pigeon and zebra finch. 

Lemnothalamic pathways & binocular vision. Binocular vision presumably evolved independently in birds and mammals, given the differences in their retinal decussation patterns.  The pattern of optic nerve decussation in birds is very different from that of the primates.  Decussation patterns relate to the extent to which the eyes are more frontally or laterally placed in the skull, and hence the degree of binocular processing that placement allows.  Most birds and reptiles have laterally-placed eyes and possess full decussation, whereby the optic nerves from both nasal and temporal hemiretina project to the contralateral side of the brain.  Primates have a partial decussation pattern, whereby only the optic nerve fibers from the nasal hemiretina cross to the contralateral side while those of the temporal hemiretina project ipsilaterally.  Partial decussation is found in most mammals, with the degree of uncrossed fibers varying from 10% in rabbit (who once were thought to lack any binocular vision) to 50% in non-human primates and man (Pettigrew, 1991).  The existence of some intermediate evolutionary step whereby  both partial decussation at the optic chiasm and a bilateral projection from the thalamus existed is unlikely if topographic relations were to be maintained in such an organism. 

Organisms exhibiting full decussation of the optic nerves from the retina may still possess binocular vision.  It was traditionally thought that partial decussation was essential for binocular vision, since it allows information from both eyes to converge on the same brain structures.  This, however, does not appear to be true due to a unique arrangement of visual projections in some birds known as a double decussation.  Predatory birds, such as owls and some diurnal raptors, tend to be those with more frontally-placed eyes and presumably larger binocular visual fields used for stereoscopic vision.  In owls, despite the initial full decussation of the optic nerves, a recrossing occurs via a second decussation from the principle thalamic optic nucleus (OPT, also called the dorsal lateral geniculate nucleus, GLd).  The fibers of the second decussation are those originating from the temporal retina, which are most involved in binocular processing (Pettigrew, 1991).  Visual thalamic neurons then project bilaterally to visual wulst (e.g., in owls, kingfishers, some nightjars, and diurnal birds of prey) where binocular integration occurs (Pettigrew, 1991).  

In birds there is a positive correlation between the extent of the lemnothalamic pathway and the development of binocular vision.  Owls have an extensively developed OPT and visual wulst (Karten et al., 1973), and binocular processing has been demonstated in owls' wulst (Pettigrew, 1979).  Lateral-eyed birds like the pigeon have a less extensive wulst and smaller binocular fields.  In pigeons, OPT receives retinal input mostly from the region for lateral viewing before sending projections to visual wulst (Güntürkün et al., 1993).  Binocular processing could still occur in the pigeon's lenmothalamic pathway, since descending projections from wulst terminate in the OPT, which then sends bilateral projections to the wulst (Güntürkün et al., 1993).  However, it is possible that collothalamic "dominance" (at least in pigeons) may relate more to feeding strategies than having lateral eyes.  A diurnal, lateral-eyed raptor, the kestral (Falco sparverius) has an over-representation of its frontal binocular visual field in the lemnothalamic pathway (Pettigrew, 1978). 

The importance of the lemnothalamic visual system in primates may relate to nocturnal habits early in mammalian evolution.  The collothalamic system of the early mammals probably remained relatively stable during their "nocturnal" period (see Bottleneck Theory, this chapter).  The topographic mapping of the optic tectum and collothalamic visual pathways should have been sufficient to serve their needs in preying on insects or perhaps small reptiles.  Other parts of the brain, including the olfactory and auditory systems may have continued to rapidly evolve.  The olfactory system already present in the mammal-like therapsid reptiles retained its importance as a means to locate prey and perhaps conspecifics for reproduction.  The auditory system, in conjunction with the rod dominated retina, could have become increasingly sophisticated, enabling efficient localization of prey under scotopic conditions.  The spectacular growth of the cerebral cortex may also have begun during this period.  This may also be the point in mammalian evolution when the importance of the lemnothalamic visual system took hold.  Connections from the retinal ganglion cells to the dorsal thalamus (LGNd in primates) were added to existing retino-collicular connections.  These retino-geniculate connections then formed increasingly complex connections with striate cortex.  These connections gained in importance in serving visual functions, functions which seem to be served by collothalamic systems in birds and reptiles. 

Support for the expansion of lemnothalamic function comes from the study of modern snakes.  Snakes are presumed to have evolved from lizards which began to live a burrowing existence and eventually "lost" their limbs.  The sensory systems of snakes appear to have been reduced during this burrowing period.  Upon re-attaining niches in the water and above ground, snakes appeared to have "rebuilt" their visual systems with a reliance on retino-geniculate connections.  The optic tectum and nucleus rotundus (collothalamic components) in snakes (e.g., the water snake Natrix) are reduced compared to other reptiles; however, the dorsal lateral geniculate nucleus is relatively large (Northcutt & Butler, 1974).  Since portions of the lemnothalamic pathways of snakes are more developed than in other reptiles, it appears that upon re-emergence from their burrowing existence, it was the lemnothalamic system which continued to evolve to serve visual functions.  Perhaps the collothalamic system was too reduced, or not neurologically plastic enough, to serve the visual processing roles which became the domain of the lemnothalamic visual pathways. 

The lemnothalamic, or geniculo-striate, system of primates is the most developed of the mammals.  In primates, approximately 90% of the retinal ganglion cells send projections to the lateral geniculate nucleus of the dorsal thalamus (LGNd), with the remainder projecting to the pretectum and superior colliculus.  Damage to the striate cortex (V1) in humans can lead to a condition known as blindsight, in which the patient is not consciously aware, but can locate a moving object in the affected visual hemifield.  This spared ability in the "blind" area is presumably the result of the processing ability of the collothalamic visual system.  In addition, the extrastriate cortex is the sight of numerous important visually-responsive areas involved in color (V4), form (IT), and motion perception (MT).  

The reliance of primates on binocular vision and stereopsis may have driven the development of the lemnothalamic visual system.  It appears that all mammals have some degree of binocular vision.  Binocular vision was probably present in the earliest mammals, and contributed to their success as a class.  Stereopsis increases the signal to noise ratio in low light conditions, serving as a camouflage breaker.   In addition, the conjugate eye movements of mammals to fixate objects of attention on the area centralis or fovea set mammals apart (Walls, 1942).  Frontal vision has also led to an increased ganglion cell density in frontally directed retina, enhancing visual acuity.  These traits are exceptionally developed in the primates.  Early fossil primates show large, frontally placed eyes and it is quite reasonable to assume they had binocular visual systems as good as those of modern prosimian primates (Pettigrew, 1991).  The binocular field is probably also related to the praxic field (i.e., the use of visually-controlled activity to manipulate objects).  The ability of the primates and their predecessors to use tools may have contributed to lemnothalamic development, as well as refinements in motor control and increases in poly-sensory cortical areas.   Binocular vision, stereopsis, and tool use must have had an extraordinary effect on visual evolution in primates. 

Collothalamic pathways in lateral-eyed birds. Evidence suggests that the lemnothalamic pathway is not as important to visual processing in lateral-eyed birds as it is in mammals.  The avian lemnothalamic pathway is comparable to the geniculo-striate mammalian pathway.  The visual wulst has both anatomical and physiological similarities to mammalian striate cortex including; laminar appearance, receipt of dorsal thalamic efferents, and cells characterized by small receptive fields.  The contribution of the lemnothalamic pathway to visual perception in the pigeon is currently unclear.  Neurobehavioral studies show surprisingly mild deficits in visual discrimination performance after lesions of the pigeon OPT and visual wulst in lateral-eyed birds.  Numerous studies employing discriminations of intensity, pattern, and color yield either no discernible deficits or only mild to moderate ones (Hodos et al., 1973; Hodos & Bonbright, 1974; Delius et al., 1984; Chaves et al., 1993).  Thus, it appears the lemnothalamic pathway is not crucial to these primary visual functions, and that the collothalamic pathway can perform simple visual discriminations without lemnothalamic processing.  It has been suggested, however, that the traditional means of testing visual discriminations with projection of the stimulus onto the response keys may affect the outcomes of these studies.  The birds are required to fixate the test stimuli with their frontal visual fields (served presumably by the tectofugal pathway), and therefore lesions of OPT or visual wulst would show little effect.  Hahmann and Güntürkün (1993) demonstrated that visual acuity of head-fixed pigeons with OPT lesions was unimpaired for frontal acuity, but subjects suffered performance deficits even for course grating patterns presented to the lateral fields.  

The collothalamic pathway appears to dominate visual functions in pigeons, and perhaps most lateral-eyed birds.  Collothalamic pathway lesions result in severe deficits in various visual discrimination tasks with brightness, color, and patterns (Hodos, 1976; Engelage & Bischof, 1993).  Such lesions can also result in significant reductions in acuity, and in thresholds for discriminating brightness and size (Hodos, Macko, &  Bessette, 1984; Hodos, Weiss, & Bessette 1986, 1988).   In particular, lesions of Rt and ectostriatum can have profound effects on visual discrimination.  Lesions of Rt have severe, detrimental effects on intensity, pattern, and color discrimination tasks (Hodos, 1969; Hodos & Karten, 1966). 

Lesions in the ectostriatum cause severe, and sometimes irreversible, deficits in visual performance for brightness, color, and pattern discriminations (Hodos et al., 1973; Hodos et al., 1984; Bessette & Hodos, 1989).  Ectostriatal lesions also affect the ability of pigeons to visually distinguish between natural concepts (e.g., food versus non-food); (Watanabe, 1991).   Lesions which involve both the lemnothalamic and the collothalamic systems result in severe deficits in discriminations of a variety of stimuli.  Such lesions produce deficits that are greater than when either pathway alone is disrupted (Hodos & Bonbright, 1974, Hodos et al., 1973; Macko & Hodos, 1984; Riley et al., 1988).  Combined lesions of the lemno- and collothalamic systems are particularly detrimental to spatial resolution tasks, producing greater deficits than birds with only collothalamic lesions (Macko & Hodos, 1984; Hodos et al., 1984). 

.The Flying Brain 

Powered flight and motion processing. The development of powered flight and the complexity of visual analysis it requires gradually changed the reptilian brain of avian ancestors to the form seen in modern birds.  Active flight requires a very high output of power (work per unit time), but is more efficient than running (but less efficient than swimming) in terms of the transport of a unit mass over a unit distance (Norber, 1985).  The demands of flight have had a tremendous impact on birds' body size, weight, skeletal morphology, feather shape and number, feeding styles, brain size, et cetera.  Powered flight has also contributed to the tremendous diversity and number of bird species, enabling them to cross natural barriers and fill almost every ecological niche on earth.   Flight requires a more specialized visual apparatus than that which birds inherited from their reptilian forebears.  

The difference between human and avian velocity thresholds may indicate that the visual system of birds is biased toward the recognition and calculations of higher velocity targets.  Thresholds for detection of velocity in pigeons, as determined by Hodos and colleagues (1975) indicated retinal velocity ranges of 4.1-6.01 deg/sec; human observers can detect retinal velocities of 3min/sec (Graham, 1968), roughly .05deg/sec.  Studies in our lab indicate that birds have more difficulty assessing the direction of moving dots on a stimulus display for "slow" speeds (16.63 cm/sec) than "fast" speeds (3.34 cm/sec); (Laverghetta and Shimizu, 1999).  On the other hand, the velocity detection data for pigeons may be underestimated, given the fact that the subjects were tested with stimuli presented in their frontal fields.  Birds (at least some species) appear to use frontal viewing for stationary stimuli but orient their lateral visual fields when viewing fast moving stimuli (Maldonado et al., 1988). 

The high proportion of directionally-selective and motion-responsive units in the collothalamic visual pathways, compared to that of reptiles, illustrates adaptations required for an air-borne lifestyle. The collothalamic visual pathway in birds exhibits high proportions of motion and directional responses at the mesencephalic, diencephalic, and telencephalic levels. Birds, with the possible exception of the diurnal primates (e.g., humans), are the vertebrates which may rely most heavily on vision to function in their environment.  The most obvious visually-dependent behavior of birds is flight.  The navigation of three dimensions where X, Y, and Z plays a role in the approach or avoidance of stimuli in the environment requires a high degree of visual processing. 

Birds also exhibit an impressive range of visually guided behaviors other than flight.  An aerial lifestyle also requires different breeding and feeding strategies.  Feeding is also dependent on visual cues in most avian species, whether a predatory raptor or a grainiverous ground-feeder.  Sophisticated visual abilities have allowed the development of good recognition of predators and conspecific mates or rivals.  These abilities have, in turn, expanded the behavioral repertoire of birds; in many species it has enabled courtship displays and the formation of pair-bonding and social hierarchies.  Given the various specializations driven by a species' individual ecology, the following ideas apply generally to most birds.  Birds brains, deriving from the basic reptilian plan, maintained a prominent collothalamic visual system.  This system became even more specialized for the demands of flight, and came to serve sophisticated visual abilities used in sexual selection, nesting, territoriality, and conspecific recognition.  

The Brains of Pterosaurs and Birds 

Pterosaurs - optic lobes. Comparing pterosaur brains with those of other reptiles and modern birds illuminates what its structure contributed to flight mechanisms, and perhaps other parts of its lifestyle which it did not share with land-bound reptiles.  A brain structure adapted to the demands of flight has occurred in evolutionary history long before the appearance of birds.  Pterosaurs, the first flying vertebrates, show adaptations undergone by a reptilian-type brain in order to achieve flight. Generally, much of the pterosaur brain can be considered as intermediate between that of reptiles and birds.  Both birds and pterosaur brains exhibit proportionally large optic lobes, small olfactory bulbs, and an expanded forebrain.  

The optic lobes in Pterosaurs were similar in position and proportions to that of modern birds.   In Pterodactylus, the optic lobes were very close to the orbit, so that the optic nerve must have been extremely short (Edinger, 1941).  Rhamphorynchus and Pteronadon endocasts which represent only the upper half of the brain, there are no discernible midbrain lobes.  These lobes have presumably descended, shifting basally from the reptilian dorsal position (Edinger, 1941).  In lateral views of Pterodactylus specimens, the lobe chamber is well-defined and relatively large.  The optic lobes probably extended over the lower half of the brain in Pterodactylus, and they were situated even further down in Rhamphorynchus.  In the later pterosaur, Pteranodon (Cretaceous), the optic lobes are confined to the basal half of the brain or lower, a position similar to that in modern birds (Edinger, 1941).  The optic lobe size indicates that they were highly developed, most likely in support of flight behaviors.   They were avian in character and very large compared to those of other reptiles of the time.  Some species of pterosaurs were small and probably tree dwelling, while some were large, gliding sea birds.  Pterosaurs exhibited many adaptations for active flight, but many larger species were probably gliders, such as the giant Quetzalcoatlus of the late Cretaceous, with an estimated wingspan of 36-39 feet (11-12 meters), close to that of a light aircraft (Wellnhofer, 1996). 

Modern birds - optic lobes. The optic tectum (TeO) in birds exhibits characteristics important for motion processing.   Approximately 90% of the retinal ganglion cell axons in lateral-eyed birds project to the highly laminated TeO (Shimizu & Karten, 1991), a structure which is likely homologous to the mammalian superior colliculus.  The retinal efferents to TeO travel the optic tract and terminate within the superficial layers of the contralateral TeO (Karten & Revzin, 1966; Karten, 1969).  Most of the deep layers of TeO possess cells which exhibit large receptive fields (70-180 degrees) and are motion-responsive and directionally selective (Frost and DiFranco, 1976).   Additionally, deeper tectal layers, like the stratum griseum centrale (SGC) have cells which demonstrate a high degree of habituation to repeated stimulus presentations (Woods and Frost, 1977).  The receptive-field size and response characteristics are traits one would expect from a motion-processing area. 

Pterosaurs - cerebellum. The highly folded cerebellum ("little brain") in all vertebrates is involved in the timing, coordination, and modulation of the motor system with incoming visual, vestibular, auditory, and somatosensory information.  The cerebellum consists of two basic divisions, the main body of the corpus cerebelli and a cerebellar auricle ("little ear"). In general the vertebrate pattern consists of input from primary or secondary nuclei of the cranial nerves and from a precerebellar nucleus, the inferior olive. The output of the cerebellar cortex is to one or more deep cerebellar nuclei which then project to the reticular formation and the nucleus ruber (red nucleus). These two areas are important sources of descending spinal pathways. A small caudal structure usually on the cerebellum's ventrolateral aspect, the flocculus, receives input from the vestibular system and sends output to motor neurons that control the muscles involved in eye movements. The cerebellum also sends output to other motor nuclei of the brainstem, portions of the thalamus involved in motor functions, and vestibular nuclei. 

The pterosaurian cerebella existed at the upper limit of reptilian brains and the lower limit of avian variations (Edinger, 1941).   The cranium of Pterodactylus elegans had two incisures which were broad and hollow, corresponding to the tuber cerebellare of modern birds (Edinger, 1941).   In the same fossil there was a "median dorsocaudal protuberance, delimited above and below by transverse furrows", which Edinger states must represent the pterodactylian cerebellum.  Unfortunately, the cerebellar folds or gyri were not impressed within the skull cavity, so much of its basic structure is speculative.   Edinger does note that descriptions of other pterosaur specimens indicate the presence of flocculi. 

Modern birds - cerebellum. Many evolutionary changes seen in the structural morphology of the cerebellum occurred with the transition of aquatic forms to the terrestrial tetropods.  With the development of locomotor abilities and behaviors for grasping, digging, et cetera, there was a concomitant lateral expansion and increase in folding.  Birds must have sophisticated cerebellar and vestibular systems to deal with the complexities of movement in three dimensions. These systems must work sufficiently to avoid obstacles, chase away encroachments on territories, or even catch meals "on the wing."  The bird's cerebellum is greatly expanded and extensively folded in comparison to modern reptiles.  Avian and mammalian cerebella are large, highly folded structures which have a very similar organization according to comparative studies. Birds have 10 large folds or lobules in their cerebella. Mammals have 10 lobules on the vermis, at the midline of the cerebellum, with many more lobules on the lateral portions.  This lateral expansion of the cerebellum in mammals could be a phylogenetically new structure, but it is more likely a variation on that seen in the other amniotes. 

Pterosaurs - forebrain. Pterosaurs (prehistoric flying reptiles) achieved an "avian-like" forebrain long before the existence of birds.  The gross morphology and relative size of the forebrain, olfactory bulb, and optic lobe associated with modern birds may have first evolved in the prehistoric flying reptiles, the pterosaurs.  In fact, Edinger (1941), in a description of a Pterodactylus specimen (Pterodactylus elegans, from about the late Jurassic), went so far as to say that the avian forebrain form was achieved by the pterodactyloidea in the Cretaceous.  The modern bird forebrain is usually wider than it is long. Early pterosaurs had brain proportions which were more reptilian, possessing a forebrain that was longer than it was wide (Wellnhofer, 1996).  This gradually changed, with the Rhamphorhynchoidea (late Triassic and Early Jurassic) exhibiting forebrain widths up to 1.5 times their length, a more "bird-like" ratio.  Pteranodon (early Cretaceous) forebrains reached similar proportions to those of modern birds.  The pterodactylus brain, as in birds, formed an arch around the back of a relatively large eye orbit.  Pterodactylusí forebrain possessed two fissures similar to that of modern birds.  Pterodactylus and similar specimens show a horizontal furrow along the middle portion of the forebrain which separates the upper and lower half of the forebrain.  This fissure would correspond to the avian valecula (Edinger, 1941).  A vertical furrow descending from the horizontal fissure indicates the presence of a vallecula Sylvii, similar to that which frontal and temporal sections of the modern birdís basal forebrain.  

Pterosaurs showed similar relative sizes between olfactory bulbs and other aspects of the forebrain.  At the rostral end of the brain chamber in the pterosaur specimen described by Edinger (1941), the tiny, tube-like chamber which housed the olfactory bulbs was preserved.  There were also no discernible olfactory tracts. By comparison, in the Tuatara (Sphenodon punctatus), a large-eyed reptile, the olfactory tracts may be almost as twice as long as the forebrain. The olfactory bulb had about 1/25th the volume of the optic lobe; the Tuatara has a 1:1 ratio of olfactory bulb to optic lobe. The ratios of olfactory bulb to forebrain and optic lobe found in Pterodactylus are similar to those found in modern birds. This strongly suggests that in pterosaurs, as in modern birds, olfaction was not as prominent a sense as was the highly developed eye and visual system.  

Modern birds - forebrain - nucleus rotundus. At the diencephalic level of the collothalamic pathway, the nucleus rotundus (Rt) of the dorsal thalamus also exhibits motion-processing regions.  Neurons of the SGC layer of the optic tectum send extensive projections to the largest single thalamic cell group in many reptiles and birds, the nucleus rotundus (Benowitz & Karten, 1976; Ulinski, 1983).  These projections are spatially organized, with specific rotundal subdivisions receiving afferents from specific tectal layers (Benowitz & Karten, 1976; Nixdorf & Bischof, 1982).  Approximately 80% of the neurons in Rt have very large receptive fields (from 100 to 175°) and are motion-sensitive (Revzin, 1970).  A series of electophysiological experiments by Wang, Jiang, and Frost (1993) indicated functional subdivisions for color, whole-field illumination levels, and motion.  The motion-sensitive neurons of Rt are found in higher proportions in the posterior Rt (Wang, Jiang, and Frost, 1993).  The posterior, in turn, exhibits cells tuned either to motion in two dimensions or that along the z-axis (motion in depth); (Wang et al., 1993).  Neurons in Rt send axons to the core ectostriatum (Karten & Hodos, 1970; Watanabe, Ito, & Ikushima, 1985).  

Modern birds - forebrain - ectostriatum. The primary telencephalic nucleus of the collothalamic pathway, the ectostriatum, shows motion-processing characteristics in birds.   The response properties of ectostriatal neurons show many similarities to those in Rt.  Ectostriatal receptive fields (150-180°) are similar to those found in Rt (100 to 175°); (Revzin, 1970).  Ectostriatal cells typically respond best to upward/downward or fore/aft movement with no selectivity for orientation (Revzin, 1970; Kimberly et al., 1971).  The ectostriatum in zebra finches also shows a high degree of motion-responsiveness.  Engelage and Bischof (1996) found large receptive fields in ectostriatum, some of which covered the entire visual field of the contralateral eye.   In addition, they performed a physiological analysis of ectostriatal neuron firing rates upon exposure to various visual stimuli.  Using bars which could be moved in directions parallel or orthogonal to their longest axis, Engelage and Bischof (1996) found about 80% of the neurons tested were motion responsive.  Of the units which reacted to bars moving parallel to their longest axis, about 70% showed orientation-selectivity.  Conversely, of those neurons reacting to bars moving orthogonal to their longest axis, only 37% showed orientation preferences.  A sensitivity, direction, and orientation analysis of ectostriatal neurons to sinusoidal gratings also revealed slight preference for forward and upward movements.  There were also units which responded vigorously to "looming" stimuli (stimuli moving towards the eye).  The orientation selectivity of neurons which responded to bars moving parallel to their longest axis could be indicative of neurons which deal with optic flow in the environment during flight.  The somewhat higher sensitivity to backward movements may also be such a specialization, and upward-movement sensitive neurons could serve as tilt detectors during flight, as previous authors have suggested (Kimberly et al., 1971; Engelage and Bischof, 1996).  In addition, units which showed high response rates to looming stimuli were similar to those of TeO and Rt.  These types of cells could theoretically be useful in calculating "time to collision" (Wang and Frost, 1992; Wang et al., 1993).  The presence of significant proportions of motion-responsive neurons in the ectostriatum, and at other levels of the collothalamic visual pathway, indicates that a significant portion of the avian visual system has developed to meet the perceptual demands inherent in the control of flight behaviors.  In addition, the ectostriatum has been shown to project to several other telencephaic areas (Ritchie, 1979; Husband & Shimizu, 1999), including the external pallium which sends significant projections to the basal ganglia (Veenman et al., 1995). 

 Next Section:  References and Links