Rothenbuhler's honeybee hygienic-behavior inheritance experiment used foul brood, an infectious disease of bee grubs, to investigate how a complex behavior can be inherited. By crossing hygienic and susceptible bee strains and examining their daughter workers, W. C. Rothenbuhler found that the behavior could be separated into stages: uncapping diseased brood cells and throwing out the larvae. The experiment became a concrete demonstration of how genetic effects can cooperate in a behavior while remaining distinct in inheritance.

Disease and Behavioral Difference

Foul brood attacks grubs in their cells. Among domestic bee breeds kept by beekeepers, some strains were more vulnerable than others, and at least part of that difference came from behavior. Hygienic strains could quickly suppress an epidemic by finding infected grubs, removing the wax cap over each cell, pulling out the larva, carrying it through the hive entrance, and discarding it on the rubbish tip. Susceptible strains failed to perform this sequence of hygienic infanticide, allowing the disease to persist.

This was not a simple, indivisible action. The worker had to locate an infected cell and then complete a succession of tasks, from opening its wax covering to disposing of the grub. That sequence made the trait useful for investigating how separate inherited influences might contribute to a behavior that, in practice, appeared to function as a coordinated whole.

Crosses and the First Generation

The genetic test required a breeding design suited to the bees' reproductive biology: worker bees do not ordinarily reproduce. Rothenbuhler therefore crossed a queen from one strain with a drone, or male, from the other, and assessed the behavior of the resulting daughter workers. The first-generation hybrid hives were all non-hygienic. At first this suggested that the hygienic parent's behavior had disappeared in the cross; the subsequent results instead showed that the relevant hygienic effects were still present but recessive.

Rothenbuhler then back-crossed the first-generation hybrids with a pure hygienic strain, again using queens and drones. The daughter hives fell into three observable groups: some showed complete hygienic behavior, some showed none, and some went only halfway. The intermediate group uncapped the wax cells of diseased grubs but did not proceed to remove and discard the larvae. This partial performance supplied a crucial clue: what looked like one behavioral capacity could be divided into two components.

Two Cooperating Genetic Effects

Rothenbuhler proposed that the sequence depended on two separate genes, one associated with uncapping and the other with throwing out the larvae. On this account, normal hygienic strains possessed both; susceptible strains carried the rival alleles at both. The intermediate bees could uncap because they possessed the uncapping effect in double dose, but lacked the effect needed to carry out the next step. The proposed division made sense of why some hives performed the first action without completing the second.

The experiment also produced a revealing test of the bees classified as entirely non-hygienic. Rothenbuhler suspected that this group might include bees with the throwing-out effect but without the uncapping effect, leaving their capacity to remove larvae concealed by the sealed cells. He removed the caps himself. Half of the apparently non-hygienic bees then displayed normal throwing-out behavior. The manual intervention exposed a behavioral capacity that had been present but could not be expressed under the ordinary conditions of the test.

What the Experiment Shows

The results illustrate why it can be meaningful to speak of a gene for a behavior without claiming that one gene single-handedly builds the entire behavior. The account does not specify every developmental cause linking genes to the workers' actions. One possible mechanism offered is that an uncapping effect might give bees a taste for infected wax, making the action rewarding and more likely to be repeated. Even if learning were part of that chain, the genetic shorthand would remain useful if, with other conditions held equal, bees possessing the relevant gene were more likely to uncap than bees carrying its alternative.

The two proposed effects also show how cooperation in an animal's behavior differs from unity in genetic transmission. The throwing-out effect is of no use without uncapping, and uncapping alone does not complete the hygienic sequence. In their functional work the effects belong together; in their passage through generations they remain, in principle, separable replicators. The experiment therefore supports a gene-centered explanation of behavioral variation without reducing the worker's complex actions to the independent work of a single gene. Its evidential force lies in the breeding outcomes, the intermediate phenotype, and the cap-removal test—not in treating an illustrative story as proof of every general claim about altruism or behavior.