Pandas in the wild are typically solitary animals. Females tend to not tolerate other females within their territory, while others interact primarily only during the short mating season, with males competing with each other over females, and males/females only making contact during mating itself. This makes the interaction between two pandas at Bifengxia all the more fascinating.
Webcam footage captured a surprisingly complex exchange between the two. We see panda 1 on the ground, panda 2 on a ledge above. Panda 2 attempts to climb down a ladder to ground level, losing its footing and getting its body completely stuck between two rungs, leaving it helpless on its back, limbs flailing in the air. Panda 1, without hesitation, then proceeds to get up and stand under the ladder and under its companion’s back for support. When that fails, panda 1 with almost human-like dexterity stands up on its hind legs, pulls the other out of the ladder, and helps it safely onto the ground. Panda 2 is clearly stunned from the ordeal and grateful for the help of its friend, as they spend several moments snuggling affectionately, then playfully rolling around together.
From this observation it is fairly clear that the animals’ interactions within their own species are far different from anything seen in the wild. Their ability to bond in this way is specific to living in captivity - as previously stated, pandas are typically solitary. Due to factors such as having few natural predators, mother-only parental care, competition over food sources, and a low energy diet and lifestyle, adults would have little benefit from face-to-face interactions in the wild.
However, certain challenges in the wild are eliminated in captivity (primarily mate, den, and food competition). This may explain the social bond, as a scenario has essentially been created in which they would have little reason to show aggression towards others in their species.
Regarding the first panda’s jump into action when help was needed - pandas have demonstrated the ability to learn based on observation both in captivity (e.g. training methods used prior to release in the wild, videos demonstrating the mating process to males), as well as in the wild (e.g. mating behaviors learned via immature males’ observation of a dominant mature male). The problem of falling through the ladder is clearly one that would not occur in nature, though panda 1 may have understood panda 2’s distress via vocal communication. Perhaps panda 1’s action came from observed behaviors in humans, other pandas, or by previous experience. Or perhaps it would be more reasonable to consider this a demonstration of trial-and-error insight learning?