Back to All Research Projects

Simple Visual Discrimination Abilities of the Giant Panda

Giant panda maternal behavior           

The primary objective of this study is to quantitatively document the giant panda mother-cub relationship. Because we hypothesize that the early social experience of the giant panda cub with its mother is crucial in determining later behavioral competence when the cubs mature, understanding the nature of this early relationship is fundamental.

This study compares the mothering styles of different females, particularly contrasting the behavior of inexperienced (i.e., primiparous) with experienced (i.e., multiparous females), females raising twins with females raising single cubs, females raising males with females raising females, and individual females over repeated births. Basic, descriptive information on giant panda maternal behavior will contribute to understanding the dynamic of this developmental process, and in identifying abnormal or deficient mothering when it does occur.

Five mothers and the 20 cubs they raised are included in this study. Four of the mothers are represented at least twice in the study, because they reared more than one litter in a particular year. We analyzed data from five mothers with cubs 0–3 months of age and 3 mothers with cubs 5–12 months of age (not all mothers reared their cubs beyond 4 months). Two of the females were primiparous and the other three were multiparous. Three females raised single cubs, and two females raised twins. The four females repeated in the study raised males one year and females another year. One female raised only female cubs.

This study is part of a long-term behavioral research collaboration with the Chengdu Research Base of Giant Panda Breeding and the Chengdu Zoo. All the subjects were housed at the Chengdu institutions. The pandas housed at Zoo Atlanta were included in this study with their mothers. One of the primary reasons these pandas were chosen for loan to Zoo Atlanta was because they were already part of ongoing studies.

It has been difficult to increase the number of subjects in this study, partly because the same females have given birth each year in Chengdu, and partly because some females reject their cubs immediately after birth, and thus cannot be included in the study because their cubs are hand-raised or cross-fostered with another mother. Rejection of cubs occurs more frequently in primiparous females, making it difficult to add inexperienced mothers to the study. This is one of many reasons giant panda births at Zoo Atlanta are important, as cubs increase our sample size for this study so that we can test all the variables of interest.

During the 0–3 month period, we found a high degree of maternal solicitude, with mothers spending 80% of their time holding cubs in the first three weeks. The amount of time mothers spent holding cubs did not fall below 50% for the first six weeks. No significant differences were found between experienced and inexperienced mothers in the amount of time they spent resting, feeding, licking, holding, or interacting with their cubs. Behavioral differences that approached significance (i.e., p = 0.08) were found between mothers that raised males compared to mothers that raised females. Mothers of males spent more time interacting with their cubs and licking the anogenital area of the cubs. Mother of females spent more time holding their cubs and licking areas of the cub other than the anogenital area.

During the 5–12 month period, play-fighting was the most common type of social interaction between mothers and cubs. We also found that mothers and cubs spent an average of 6% of their time engaged in reciprocal play-fighting (i.e., rough and tumble play) when cubs were 6–12 months old. In contrast, cubs spent less than 3% of their time nursing, which was the other prevalent mother-cub interaction. No agonistic behavior was observed between mothers and cubs. No significant differences were found in the amount of time mothers engaged in reciprocal play-fighting with male and female cubs. Comparison of mothers raising twins with mothers raising singletons resulted in no significant differences for reciprocal play-fight, non-reciprocal play-fight, feed, self-groom, locomote or rest. The sample size was too small to test for differences between experienced and inexperienced mothers.

Our results demonstrate that intensive maternal care in the early weeks of life is necessary for extremely altricial giant panda cubs. Mothers were found to care differently for males than for females during the early part of development (0–3 months), but not after 4 months of age, when cubs become mobile and eliminate waste without assistance from the mother. This indicates there might be a sensitive period early in giant panda development during which differential treatment of the sexes by the mother is important for species-typical development. We also found that mother-cub social interaction remained high beyond 6 months of age, with play-fighting being the most prevalent social interaction. Although we have found that male cubs direct more non-reciprocal play at mothers (Snyder et al., 2003), it seems that mothers respond equally to play invitations from both male and female cubs.

These results, combined with our finding that mothers frequently initiate play-fighting, indicate that giant panda mothers take a more active role in play than has been reported for mothers of other species. These findings suggest that cubs removed from their mothers at an early age will be deprived of important social experiences, which may be necessary for the development of species-typical social behavior, including reproductive behavior.

Simple visual discrimination abilities of the giant panda

Recent research on giant panda chemical communication has greatly increased our understanding of this aspect of giant panda sensory perception. Other sensory systems have not yet been investigated, and almost nothing is known about the visual abilities of pandas. Although giant pandas do not seem to have need for a highly developed visual system, most diurnal mammals, including other bear species, are postulated to have at least some level of color vision. The eye of the giant panda contains rods and cones, with the rods outnumbering cones. This suggests, but does not guarantee, that giant pandas possess acute night vision and are capable of color vision. Experimental testing of black bears has shown that these animals are able to discriminate green and blue. The purpose of this study was to use methodology similar to that used to test black bears to determine if giant pandas were capable of discriminating colors.

The two giant pandas on loan to Zoo Atlanta were study subjects. This study consisted of two phases. The first phase was conducted to gather empirical evidence that giant pandas can be trained using operant conditioning to respond to the test apparatus, and that they can learn visual discrimination. The second phase of this study was used to determine if giant pandas are able to distinguish colors. In phase one, the stimuli were the 18 shades of gray, including white and black. In phase two, the stimuli were five shades of each color, green, blue, and red, which were compared with the gray stimuli.

Both giant pandas were able to discriminate the white stimulus from two gray stimuli (male in 317 trials, female in 250 trials), and the black stimulus from one gray stimulus (male in 824 trials, female in 822 trials). In addition, both pandas were able to discriminate green-vs.-gray stimuli (male 759 trials, female 860 trials) and red-vs.-gray stimuli (male in 120 trials, female in 560). The female was also able to discriminate blue-vs.-gray stimuli (in 405 trials). The male was not tested on this color. For most of the testing, the male had higher than chance performance throughout, whereas the female’s performance was more variable and often dipped below chance. Neither subject consistently learned faster than the other, but the male showed some tendency of the “learning to learn phenomenon,” as evidenced by his extremely rapid discrimination of red vs. two grays.

Giant pandas were able to discriminate shades of red, green and blue stimuli from numerous shades of gray. Because brightness was eliminated as a cue, these discriminations can be attributed to color vision. Therefore, the present study provides experimental evidence that giant pandas’ visual abilities, including color vision, are comparable to those of other bears and other non-nocturnal carnivores. Giant pandas, like other carnivores that have been studied, are most likely dichromats. The discovery of a neutral point, which is the identifying trait of dichromatic vision, would confirm this claim. In addition, further behavioral tests should be performed to identify if giant pandas fail to discriminate bluish-green stimuli from grays in order to fully explore the limits of giant panda color vision. The testing procedure developed in this study also allows for expansion into olfactory and auditory modalities. It has been found for many species that performance on discrimination tasks is confounded by their dominant sensory modality. Therefore, using this procedure to measure other sensory modalities in giant pandas could help identify the giant panda’s dominant modality.