A serial-lesion technique was used to investigate interactions in visual processing between telencephalic components of the pigeon visual system. Pigeons were trained to discriminate pairs of stimuli that differed in color, intensity or pattern. After mastering the discrimination tasks, they were assigned to one of three groups. The first group (WI-EII) received lesions of the visual Wulst and were retested. After the discrimination tasks were again mastered, a second set of lesions was made, this time in the ectostriatum. The birds were tested once again after the second surgery. The second group (EI-WII), underwent the same sequence of events except that the order of the lesions was reversed. In the third group (E + W), lesions of both the visual Wulst and ectostriatum were made in a single operation, followed by retesting. The performance after the first lesion of the subjects in each of the two-stage lesion groups was typical of performance after such lesions; i.e. the birds with visual-Wulst lesions showed little or no impairment on any of the tasks, whereas the pigeons with ectostriatum lesions showed considerable deficits in intensity and pattern discrimination, which diminished after prolonged retraining. In contrast, the pigeons in the one-stage group (E + W) showed profound deficits that appeared to be permanent. The performance after the second operation of the WI-EII group was the same as that of pigeons with lesions of ectostriatum alone; i.e. destruction of ectostriatum first or second resulted in the same duration of impairment. The performance of the EI-WII group after its visual Wulst lesion, however, was similar to that observed in the E + W group. The results are interpreted as a reflection of parallel processing within the avian visual system; i.e. the presence of an intact tectofugal pathway may mask the effects of thalamofugal pathway interruption.

Pigeons were trained to discriminate visual stimuli that differed in intensity, color, or pattern, after which lesions were made in the ectostriatum, either the core region (EC), the belt region (EB), or both (EC + EB). Control lesions were made in the lateral neostriatum. After surgery, the birds were again trained to discriminate the stimuli. On the basis of a quantitative analysis of the lesion reconstructions, each pigeon was assigned to one of five groups: EC, EB, moderate (EC + EB), slight (EC + EB), or control. Postoperative savings scores, based on the number of preoperative and postoperative sessions required to discriminate the stimuli, were calculated. An analysis of variance indicated a significant effect of lesion location on the extent of postoperative savings or loss. Only the moderate (EC + EB) group was significantly different from the neostriatum controls. A multiple regression analysis of performance on the individual discrimination tasks indicated that the extent of damage to EC predicted the initial performance deficit on the intensity and one of the two pattern discrimination tasks (vertical versus horizontal bar). The color discrimination deficit, which was modest in extent and of short duration, was predicted only by the damage to (EC + EB) that was at least moderate in extent (20–40% of the volume of (EC + EB)).

The morphological development of the tectofugal pathway in the zebra finch has recently been described in a series of studies from our laboratory. No data are currently available on the development of visual responsiveness in this pathway. We therefore investigated the development of visually evoked potentials (VEPs) in the ectostriatum, the telencephalic target area of the tectofugal pathway. Contralateral VEPs could already be recorded in 20-day-old birds, whereas ipsilateral VEPs could first be recorded in 40-day-old birds. The latencies of contralateral VEPs decrease to adult values between 20 and 40 days of age, probably due to an increase in the myelination of afferent fibers. The amplitudes of the contralateral VEPs increase continuously from day 20 to day 60; however, between 60 and 80 days of age the responses diminish substantially (−60%). Thus, contralateral VEPs in 80-day-old birds are not significantly different from those in 20-day-old birds. Thereafter the responses recover and reach their final amplitude values at about 150 days of age. The relationship of these results to morphological studies and possible mechanisms which may cause the double-peaked development of visual-evoked potentials in the ectostriatum are discussed.

Pigeons were trained on four different visual discrimination tasks: (1) concept of natural stimuli (food vs. non-food object discrimination); (2) arbitrary classification of natural stimuli (pseudoconcept); (3) concept of artificial stimuli (triangles generated by computer graphics); and (4) discrimination of one pair of artificial stimuli. Then, lesions of the ectostriatum were carried out. The ectostriatal lesions impaired the arbitrary classification of natural stimuli and the concept of artificial pattern but did not impair the natural concept or the simple discrimination of fixed two stimuli. Lesions in the neostriatum did not cause deficits in any discrimination task. The birds had to learn individual stimuli for the arbitrary classification of stimuli and the stimulus generalization test after the artificial pattern concept discrimination indicated that the pigeons formed a concept more complicated than “triangle” in human language. These results suggest that the ectostriatum plays a role in task discrimination that requires much visual processing to classify stimuli.