Abstract
The cyclic ratio (CR) schedule is a behavioural assay developed to study feeding in rats, in which the number of operant responses required to obtain food reward (the ratio requirement) increases and then decreases in a repeating cycle. In a recent study, we used the CR schedule with European starlings (Sturnus vulgaris) to investigate the effects of an early-life manipulation on adult feeding behaviour. As this was the first time the CR schedule had been used with any bird species, a more in-depth evaluation is warranted. Here, we performed a fuller CR experiment with the same birds as the prior study, a year later. First, we examine the individual consistency of feeding behaviour between experimental sessions and also between CR schedules comprising different ratio requirement progressions. We found that between-session consistency was poor to moderate, and that a geometric ratio progression provided greater between-session consistency than an arithmetic ratio progression. Second, we tried to replicate some of the canonical findings from rats working on CR schedules. In contrast to findings from rats, we found that defence of feeding rates did not increase when starlings were acutely food deprived. However, as in rats, we found that the post-reinforcement pause increased linearly with the upcoming ratio requirement, suggesting that starlings were able to learn the cyclic nature of the schedule. Third, we compared the results from the present study concerning the impacts of our early-life treatment with those from our earlier study. We found that the majority of our previous findings were replicated in the same individuals one year on, reinforcing our previous conclusion that the early-life manipulation had canalised our birds into two groups with different patterns of feeding rate defence.