Wilkinson’s vampire bat blood-sharing observations turn an apparently simple act of feeding into a test of how cooperation might persist between unrelated animals. When a bat found a meal, it could share blood with a less fortunate roostmate; the evidence included both relatives and unrelated bats, with the latter tending to be frequent companions. The observations therefore offered a possible instance of reciprocal altruism: help given now might be returned after fortunes changed, provided individuals met again and could distinguish one another.

The problem of unequal nightly fortunes

Vampire bats feed at night, and finding a meal is difficult. A successful bat may take in more blood than it immediately needs, while another returns to the roost without having fed. On a later night, the outcomes may reverse. This alternation between temporary plenty and need creates the circumstances for sharing: a donor that has succeeded can help a hungry companion, and the donor may later face hunger itself.

The pattern matters because the gift is not automatically explained by close genetic relationship. Among 110 instances of regurgitation witnessed by Wilkinson, 77 could readily be understood as mothers feeding their offspring, and many other cases involved genetic relatives. Some donations, however, were between unrelated bats. The account treats these cases as the particular puzzle: relatedness could not by itself explain why a donor helped, so repeated association and possible repayment became relevant.

From observation to an exchange model

The book places the observations alongside a question about unpaid blood donation among humans, but resists explaining human donors through reciprocity or concealed self-interest in any straightforward way. Vampire bat sharing, by contrast, is presented as a better fit for the Axelrod model. The analogy is not that the bats consciously calculate a bargain; it is that the conditions of their lives can make repeated cooperation advantageous.

The proposed exchange is asymmetric in its immediate costs and benefits. A bat giving up blood risks reducing its own prospects, but the same amount can matter more to a severely hungry recipient than to a less hungry donor. Wilkinson estimated how quickly bats at different levels of starvation would lose weight and used the resulting survival times to express blood in hours of life. The calculation suggested that a given quantity could extend the life of a highly starved recipient more than it reduced the donor’s prospects. On that account, giving could be a modest cost to the successful bat and a substantial benefit to the hungry one—the payoff structure expected in a Prisoner’s Dilemma.

This proposed economy also gives defection a specific meaning. On a night of plenty, a donor might benefit slightly by withholding blood; on a night of need, that same bat could benefit greatly from receiving help. Such withholding would be consequential only if bats had opportunities to meet again and if prior conduct could affect later treatment. The recurring contact among roostmates supplies the first condition; recognition and response are needed for the second.

Repeated contact and recognition

The unrelated bats involved in sharing tended to be frequent roostmates, and therefore had repeated opportunities to interact. The book treats this recurrence as a prerequisite for an iterated Prisoner’s Dilemma: a single encounter would offer no future return for cooperation, whereas repeated association permits present generosity and later response to be linked. Sharing among roostmates thus appears not as an isolated event but as behavior within an ongoing social setting.

Wilkinson tested whether bats could recognize individuals using captive animals drawn from two groups originally taken from caves many miles apart. On each trial, one bat was removed for a night and left hungry while the others were fed; it was then returned to the roost, and the experimenter observed which bats gave it food. The procedure was repeated with different bats taking turns as the hungry individual. Because the population combined bats from separate caves, the experiment could distinguish donations from familiar companions from donations to newcomers.

Thirteen donations were observed. In twelve cases, the donor was an old companion from the hungry bat’s original cave; in one, the donor was a new companion from the other cave. The account calculates that the odds of this distribution arising by chance were less than one in 500, and concludes that the bats were biased toward feeding old companions rather than strangers. The result supports individual recognition and the importance of prior association, though it does not by itself establish that every donation was repayment for a previous one.

Reciprocal altruism and its limits

The observations bring together several conditions for reciprocal cooperation: a donor sometimes has a surplus, a hungry recipient can gain substantially from a donation, the same individuals may encounter one another repeatedly, and bats can favor familiar companions. In the book’s interpretation, this combination makes Tit for Tat plausible. A bat could cooperate by sharing when fortunate and, in later encounters, respond to its partner’s conduct; the strategy’s significance lies in the repeated structure of the relationship, not in any claim that bats consciously follow a rule.

The evidence also preserves an important distinction between kin-directed care and sharing beyond kin. Most of the 110 recorded donations were compatible with feeding offspring or other relatives, so those cases do not demonstrate reciprocity among unrelated partners. The smaller set involving unrelated bats is the sharper test, and the experiment on familiar companions bears on whether partners can be recognized. Together, the field observations and captive test make reciprocal altruism a credible interpretation of some blood-sharing, while leaving the actual history and terms of any particular exchange unstated.

The episode’s significance is therefore evidential as well as theoretical. A striking act of generosity is not treated as proof of disinterested motive, nor is it reduced to kinship when the recipients include unrelated companions. Instead, the account examines the conditions under which help can be sustained: uneven luck, unequal value of the same resource, repeated encounters, and discrimination between familiar partners and strangers. Wilkinson’s observations make vampire bat sharing a concrete candidate for cooperation maintained through reciprocity, while the book presents that conclusion as an evolutionary interpretation of the observed pattern.