EXPERIMENTAL-EVIDENCE FOR THE MAINTENANCE OF FORAGING SPECIALIZATIONSBY FREQUENCY-DEPENDENT CHOICE IN FLOCKS OF SPICE FINCHES

Three hypotheses can account for consistent foraging specializations a cross individuals: (1) food source variation, (2) phenotypic differenc es and (3) frequency-dependent choice. We examined which explanation w as more relevant for flocks of spice finches (Lonchura punctulata) exp loiting environments where food could be obtained using two different skills. The food source variation hypothesis predicts that individuals will specialize when the use of two skills is more costly than specia list foraging. The phenotypic differences hypothesis proposes that ind ividuals differ in their ability to use each foraging skill and specia lize on the most profitable one. The pattern of specialization is expe cted to be stable although the number of individuals that use a given skill depends on the phenotypic composition of the flock. The frequenc y-dependent choice hypothesis also proposes that individuals specializ e on the most profitable skill. However, the profitability of each alt ernative decreases as the number of foragers that use it increases. Th e hypothesis thus predicts that an equal number of phenotypically iden tical foragers would gradually specialize on each skill when initially given two equally profitable alternatives. At equilibrium, individual payoffs should be independent of the pattern of specialization. Indiv iduals in flocks of spice finches adjusted their use of the two skills and two birds in each flock specialized on a different skill. We fail ed to find a significant relationship between specialization and intak e rate at the end of testing. The results are compatible with both a v ariant of the food source variation hypothesis, where individuals expe rience the cost of generalist foraging differently, and the frequency- dependent choice hypothesis.