Self-feeding device is extensively used in aquaculture farms, but for salmonids the individual feeding behavior has seldom been continuously observed. In this article, the individual self-feeding behavior of 10 rainbow trout was continuously monitored with a PIT tag record for 50 days with three replicates. The fish fell into three categories according to their feeding behavior, i.e. high triggering fish (trigger behavior more than 25% of the group, HT), low triggering fish (1%–25%, LT) and zero triggering fish (less than 1%). The results showed that in a group of 10 individual 1–2 HT fish accounted for most of the self-feeding behavior (78.19%–89.14%), which was far more than they could consume. The trigger frequency of the fish was significantly correlated with the initial body weight (P <0.01), however, no significant difference in growth rate among the HT, LT, and ZT fish was observed (P >0.05). Cosinor analysis showed that the two HT fish in the same group had similar acrophase. Though some of the HT fish could be active for 50 d, there were also HT fish decreased triggering behavior around 40 d and the high trigger status was then replaced by other fish, which was first discovered in salimonds. Interestingly, the growth of the group was not affected by the alternation triggering fish. These results provide evidence that in the self-feeding system the HT fish didn’t gain much advantage by their frequent self-feeding behavior, and high trigger status of the HT fish is not only an individual character but also driven by the demand of the group. In the self-feeding system, the critical individual should be closely monitored.
