Newly hatched chicks spontaneously peck at conspicuous objects, and soon learn to discriminate between edible food particles and inedible objects. To examine whether this discrimination is based on a chick's ability to memorize objects by shape cues, we analyzed the pecking behavior. One- to 3-day old quail chicks (Coturnix japonica) were presented with dry objects of different shapes (ball, disk, triangle and T-shape) of similar size (4 mm) and color (green). Habituation occurred after repeated presentation of any one of these objects (duration: 30 sec; interval: 4 min). When chicks showed significantly more pecks at a novel object (dishabituation), we assumed that chicks had memorized the habituated shapes and distinguished the novel object. Chicks did not show dishabituation between a ball and a disk. On the other hand, chicks discriminated a triangle or T-shape from the memorized image of disk, but did not memorize either triangle or T-shape by its shape. Similarly, chicks did not memorize the size of disks as a reference for subsequent pecking behavior. Chicks proved to have a limited ability to memorize shape and size cues for selective pecking behavior, in strong contrast to their accurate memorization of colors.
INTRODUCTION
Newly hatched chicks of precocial birds have an innate tendency to peck at small conspicuous objects such as bright and colorful beads. When the object is edible or positively rewarding, the chick will subsequently show enhanced pecking that is often followed by biting and swallowing. When the object is not edible or not rewarding at all, the chick will become habituated to the object and gradually cease pecking. When the object tastes bitter, on the other hand, the chick will learn to avoid similar objects after a single experience of the bad-tasting bead. A variety of learning tasks have been developed based on these tendencies of chicks, i.e., one-trial passive avoidance task (Cherkin 1969, Rose SPR 1991), visual habituation (Aoki et al. 2000, Sakai et al. 2000), and rewarded GO / NO-GO task (Yanagihara et al. 2001).
Color is thought to be one of the dominant cues for the visual categorization of objects. Osorio and his colleagues (1999) have shown that chicks form accurate memories of color. In the passive avoidance learning paradigm, chicks precisely memorized the coloration of a bitter tasting bead even without any differential pre-training experiences (Aoki et al. 2000). Similar precision in the memorization of color has been documented with a water-reinforced GO pecking task, where chicks learned the cues of appetizing objects (Izawa et al. 2001).
Other non-color cues have also been suggested in chicks. A localized brain lesion is reported to cause color blindness, while preserving chicks' ability to avoid the bitter-tasting bead (Patterson and Rose 1992). The ablated region (intermediate medial hyperstriatum ventrale, or IMHV) in the dorsal telencephalon is thought to be critical in both passive avoidance task (Rose SPR 1991) and imprinting (Horn 1985, Horn 1998)). Patterson and Rose (1992) have argued that the IMHV is responsible specifically for the visual association of colors and selective avoidance, whereas the ablated chicks could avoid the bitter-tasting object on the basis of non-color cues such as shape or size. However, so far there has been no evidence directly supporting that chicks memorize shape cues for the visual categorization of objects. The purpose of the present study was to determine whether and how shapes could serve as significant visual cues for chicks to recognize objects. The results of this study support our preliminary data (Sakai et al. 2000), which suggested that chicks are innately predisposed to memorize a limited class of shapes such as balls or disks, and discriminate other shapes in reference to these memorized circular images.
MATERIALS AND METHODS
Subjects and housing conditions
Newly hatched quail chicks (Coturmix japonica) were used. Fertilized eggs were obtained from a commercial supplier, and incubated in a dark incubator kept at 37.7°C. Chicks were left in the incubator for about 24 hr after hatching. Afterwards, healthy chicks were labeled with leg rings and randomly paired in small transparent plastic cages (13×9 cm2, 9 cm high). The cages were kept in a breeder that was illuminated by dim white light (35–80 lux, 12 L: 12 D), highly moisturized and kept at a controlled temperature between 31–35°C. The chicks were given a petri dish containing a water-soaked sponge and allowed to drink freely. From 48 hr post-hatch, chicks were also given quail mash food. In this study, experiments were performed at 24–72 hr post-hacth.
Visual objects for discrimination
Six objects were used, i.e., 4 of different shapes and 2 of different sizes (Fig. 1C). Objects were presented in the orientation shown in Fig. 1C, and not rotated along the bar axis. The surface of objects was painted green (type X-28 acrylic paint; Tamiya Co., Japan), and the objects were glued to transparent plastic rods (1.5 mm in diameter). The ball was a 3-dimensional plastic sphere. A disk, triangle and T-shape were made of 1.0 mm thick acrylic film. When compared among these shapes, number of pecks did not significantly differ upon the first encounter (1st trials in Fig. 3A, B, and C), thus suggesting that chicks did not have an innate biased to peck at specific shapes. Disks of two other sizes, one smaller (× 1/2 disk, or 2.8 mm in diameter) and another larger (× 2 disk, or 5.6 mm in diameter) were also made of the same film. One of these objects was presented through a hole on the wall. The orientation of the objects was fixed as illustrated in Fig. 1C, and not rotated along the rod axis. Note that the ball gives an invariant disk image on the retina irrespective of the chick's viewpoint, while the image of the other objects would vary depending on the angle from which they are viewed. The surface area of these objects was as follows; disk: 12.6 mm2, triangle: 6.9 mm2, T-shape: 7.0 mm2, × 1/2 disk: 6.2 mm2, and × 2 disk: 24.6 mm2, respectively.
General procedure of tests
The subject chick was always paired with another “soother” chick to suppress distress calling by the subject (Hayashi et al. 2001). During experiments, the pair was placed in a chamber (15×11 cm2, 18.5 cm high) within a cabinet (Fig. 1A). One sidewall of the cabinet was equipped with a transparent acrylic window. The inside of the cabinet was illuminated (ca. 1,080 lux) and thermo-controlled at 30±1°C (Fig. 1C). Before each session, the chicks were given 3 min in the box for acclimation. The subject was then separated from the “soother” chick by a transparent acrylic plate, and left for another 3 min. The plate was used in order to suppress social facilitation of pecking by the “soother”. The subject chick was thereafter examined for its preference in pecking behavior. Test objects were presented at 4 min intervals. The object was left for up to 4 min until the chick turned to or approached the object; if a chick failed to turn to / approach the object, it was excluded. The number of pecks was counted during a 30 sec period starting from the first turn or approach. Objects were manipulated by an experimenter, and presented by hand through a small hole (7.0 mm in diameter) in the wall of the chamber located 3.0 cm above the floor. Dry objects were used in all experiments. The objects were presented one at a time. The object was in a fixed orientation, and not rotated along the rod axis. The behavior of the chicks was observed through a Plexiglas one-way window of the chamber. Mann-Whitney's U-test was used to compare two independent sets of data. Sign test was used to compare the actual number of pecks (the 4th [Fig. 2, 3, 5], or the 7th [Fig. 4]) with expected number (average of the 3rd and the 5th [Fig. 2, 3, 5], or average of the 6th and the 8th [Fig. 4]). The level of significance was set at 0.05 or below.
Fig. 2
With visual cues alone, chicks did not discriminate between the 3-dimensional ball and the 2-dimensional disk of the same size. The number of pecks (ordinate; mean±S.E.M.) during 30 sec was plotted against the trial number (abscissa). The presentation procedure is illustrated in the inset diagrams. Filled circles denote ball, and open circles disk. The n indicates the number of chick pairs used in each group. A: Without the transparent side window, chicks showed significant dishabituation to the disk after being habituated to the ball. The number of pecks in the 4th trial was significantly greater than the expected value (average of the 3rd and the 5th) with p=0.038. B: With the side window, however, chicks did not show dishabituated pecking in either group. No significant difference was found compared with the expected value (i.e., average of the 3rd and the 5th trials) (ns); Signed rank test (p>0.05). No significant difference was found between group 2 and 3 (NS); Mann-Whitney's U-test (p>0.05). Results of similar statistical tests are also shown in the following figures.
Fig. 3
Chicks discriminated the triangle and T-shape from the disk. A: After being habituated to the disk, chicks showed significant dishabituation to the T-shape (group 4). In group 5, where an opposite combination was examined, chicks failed to show dishabituation to the disk after the T-shape (ns). Between the two groups, the number of pecks did not differ between corresponding trials (NS; 1st, 2nd, 3rd, and 5th). B: Similar results were obtained between the disk and triangle, although the degree of dishabituation was weaker for the triangle than the T-shape shown in A. C: Chicks did not show dishabituation between the triangle and T-shape (ns).
Fig. 4
Effects of prior experiences of the disk on the discrimination between the triangle and T-shape. A: After habituation to the triangle in the 4th to the 6th trials, chicks showed a significant dishabituation to the T-shape in the 7th trial (p=0.019). B: After habituation to the T-shape, chicks failed to show a dishabituation to the triangle in the 7th trial (ns).
Fig. 5
Chicks discriminated among disks of different sizes, though loosely. A: Dishabituation did not occur between the disk and × 2 disk. B: There was also no dishabituation between the × and × 1/2 disk. C: After being habituated to the × 2 disk, chicks showed dishabituated pecking at the × 1/2 disk. However, chicks did not show dishabituation to the × 2 disk after the × 1/2 disk. Note that smaller objects tended to induce more pecks in the first trial. Comparisons between groups by Mann-Whitney's U-test; *: p < 0.05, **: p < 0.01, NS: p ≥ 0.05.
RESULTS
Chicks did not visually discriminate between ball and disk
What mental images could chicks make upon pecking at a 3-dimensional ball? On the retina, the ball may give a disk-shaped image similar to the 2-dimensional disk. Otherwise, chicks may discriminate between these two objects based on differences such as surface shading and texture. Another possibility is that chicks rely on the tactile sensation caused by direct touch of the beak, and distinguish objects of different physical properties. The following three experimental groups were designed in order to differentiate among these possibilities.
Chick pairs were randomly assigned into 3 groups, and subject chicks were given 5 successive presentations. The same object was presented in the 1st to the 5th trials, except for the 4th trial when a different object was presented. These 3 groups are as follows:
Group 1: ball - ball - ball - disk - ball (without the side window)
Group 2: ball - ball - ball - disk - ball (with the side window)
Group 3: disk - disk - disk - ball - disk (with the side window)
Without the side window (group 1), chicks habituated to a ball from the 1st to the 3rd trials, and showed a dishabituated pecking at a disk (the 4th trial in Fig. 2A), suggesting that chicks discriminate between ball and disc. With the side window, therefore without any specific tactile cues available, chicks showed a similarly habituated pecking from the 1st to the 3rd trials. However, they did not show a dishabituation to the novel object presented in the 4th trial (group 2 and 3) (Fig. 2B). A comparison of groups 2 and 3 revealed no significant difference in the number of pecks between corresponding trials (1st, 2nd, 3rd, and 5th), suggesting no biased preference between the ball and disk. Chicks might have ignored 3-dimensional visual cues such as binocular disparity and surface texture of the objects. The 2-dimensional visual cue of the outline may thus be critical for visual discrimination of objects. In the next experiment, we examined whether chicks could discriminate between a circular object (disk) and other objects with distinct outlines (triangle and T-shape).
Chicks discriminated triangle and T-shape from a memorized image of disk
In the next step, we examined the discrimination between disk and T-shape (Fig. 3A) in the following 2 groups:
In this and the following experiments, the subject chick was separated from the object by a side window. After being habituated to a disk, chicks showed a clear dishabituation to a T-shape (group 4). However, in group 5 where an opposite combination was examined, chicks did not show dishabituation to a disk. No biased preferences were found between the two objects in the 1st to the 3rd, and the 5th trials. The apparent asymmetry between these objects suggests that chicks can memorize the disk shape and distinguish the T-shape, whereas they can not memorize the T-shape object by its shape.Similar results were obtained between disk and triangle (Fig. 3B) in 2 other experimental groups:
Group 6: disk - disk - disk - triangle - disk
Group 7: triangle - triangle - triangle - disk - triangle
We compared the triangle and T-shape (Fig. 3C) directly in the following 2 groups:
Dishabituation did not occur in either group. Under these conditions, chicks did not discriminate between these two objects, both of which were composed of straight lines and edges. How do chicks distinguish between the triangle and the T-shape, if chicks were tested after they had successfully discriminated these objects from the disk? Once they have learned to discriminate objects by shape, chicks might pay specific attention to the shape cue. In the next experiment, we thus examined the effects of preceding experiences.Chicks discriminated T-shape from triangle, once they were habituated to disk
Chicks were initially habituated to a disk in 3 successive trials, and then tested according to the same procedures as in groups 8 and 9. Eight successive presentations were thus given. These groups are as follows:
Group 10: disk - disk - disk - triangle - triangle - triangle - T - triangle
Group 11: disk - disk - disk - T - T - T - triangle - T
Chicks discriminated objects of different sizes, though they did not memorize the size as a significant cue
Both the triangle and T-shape had a smaller surface area than the disk, and chicks might have discriminated between these objects not by their shapes, but by the different sizes. As the final step in this series of experiments, we examined whether a size cue could be critical. To avoid complications, disks of 3 different sizes were used in the following 6 experimental groups:
Group 12: disk - disk - disk - × 2 disk - disk
Group 13: × 2 disk - × 2 disk - × 2 disk - disk - × 2 disk
Group 14: × 1/2 disk - × 1/2 disk - × 1/2 disk - disk- × 1/2 disk
Group 15: disk - disk - disk - × 1/2 disk - disk
Group 16: × 1/2 disk - × 1/2 disk - × 1/2 disk - × 2 disk-× 1/2 disk
Group 17: × 2 disk - × 2 disk - × 2 disk - 1/2 disk - × 2 disk
DISCUSSION
The results obtained in this study are summarized in Fig. 6. Using visual cues alone, chicks did not discriminate between the ball (3-D object) and disk (2-D object), both of which should give rise to a similar circular image on the retina (Fig. 2). Chicks could distinguish the other shapes (triangle and T-shape) from the memorized image of the disk (Fig. 3A, B). Most probably, chicks were concerned with whether the outline of the object was composed of curves (disk) or lines and edges (triangle and T-shape). When the triangle and T-shape were compared directly, however, chicks did not show dishabituated pecking (Fig. 3C). On the other hand, when tested after being habituated to the disk, chicks showed dishabituated pecking at the T-shape after the triangle (Fig. 4A), but not with the reverse order (Fig. 4B). Finally, chicks distinctively pecked at disks of different sizes (Fig. 5). All the results obtained in the present study basically support the findings of the previous report (Sakai et al. 2000), indicating that chicks' ability to recognize shapes is limited in comparison with their ability to recognize colors (Aoki et al. 2000). When discriminating colors, the degree of habituation accurately reflected the physical differences among colors. Furthermore, chicks discriminated green from yellow, and similarly yellow from green; color discrimination proved symmetric. As far as the pecking behavior is concerned, the color cues are much more critical than the shape cues.
Fig. 6
Summary of the present experiments, schematically illustrating the visual categorization of shapes in quail chicks. Note the asymmetric generalization between the categories “curves” and “lines & edges.” See text for further explanations.
Did chicks directly discriminate between the triangle and T-shape?
As shown in Fig. 4A, the preceding experiences of pecking at the disk affected the following discrimination between the T-shape and triangle. Without such a preceding experience, chicks did not discriminate between these shapes (Fig. 3C). The discrepancy between these two experiments (groups 8 and 10) may be interpreted in either one of the following ways. (1) After chicks learned to discriminate the triangle from the memorized image of the disk, they became more attentive to the difference in shape cues, and thus could directly discriminate the T-shape from the memorized image of the triangle. Alternatively, (2) chicks could memorize only the shape of the disk, and compared the two shapes (triangle and T-shape) in reference to the memorized image of the disk. To the chicks' eyes, the T-shape might have seemed more distinctive from disk than the triangle did. Actually, the stronger dishabituation appeared with the T-shape than the triangle (4th trials in group 4 vs. group 6; 4th trials in group 11 vs. group 10), thus favoring the latter interpretation. On the other hand, we do not have any lines of evidence, independently supporting chick's ability to memorize the triangle shape.
Studies of filial imprinting report that chicks can memorize and discriminate between different shapes such as a box and cylinder (Horn 1985). It has also been shown that domestic chicks can be imprinted with a cardboard triangle (Regolin and Vallortigara 1995). Definitely, chicks have the capacity to memorize complex shapes, and results obtained using the imprinting paradigms are contradictory to the present results. The difference in object sizes might explain the discrepancy; the objects used in imprinting (5–20 cm) are much bigger than the objects used in this study (2–6 mm), so that chicks could memorize the shape much more easily. Alternatively, a difference in the behavioral paradigm might explain the discrepancy; chicks might form a detailed shape image of the imprinting object, but simply do not care about the shape of objects they peck at. Otherwise, the difference in the training time might explain the different results. In imprinting, chicks are usually exposed to objects for a period of 1 to 2 hr. On the other hand, chicks might stare at objects they peck at for only a few seconds to a minute.
Did chicks directly discriminate between different sizes of disk?
As shown in group 17 (Fig. 5C), chicks showed a dishabituated pecking at the × 2 disk after being habituated to the × 1/2 disk; the size ratio was 1:4. It is also evident that chicks have an innate preference for smaller objects over larger ones (Fig. 5B, 5C). These results can be interpreted in the following two ways. (1) Chicks memorized the size of object they pecked, and discriminated among objects of different size based on the memorized size. Or, (2) Chicks behaved differently when presented with different size objects, simply because they preferred smaller objects. If the former is correct, we could expect a dishabituated pecking also at a × 2 disk after being habituated to a × 1/2 disk. However, this was not the case, and chicks failed to show a dishabituation (group 16). Although chicks may have an internal (mental) scale for measuring the size of objects, chicks do not memorize the size they measured.
Acknowledgments
We would like to thank Dr. P. Kabai (St. Istvan University, Hungary) for his critical comments. This study was supported by Grants-in-Aid for Scientific Research from the Japanese Ministry of Education, Culture, Sports, Science and Technology (#13640677), and from Takeda Science Foundation.
