Parental care involves elaborate behavioural interactions between parents and their offspring, with offspring stimulating their parents via begging to provision resources. Thus, begging has direct fitness benefits as it enhances offspring growth and survival. It is nevertheless subject to a complex evolutionary trajectory, because begging may serve as a means for the offspring to manipulate parents in the context of evolutionary conflicts of interest. Furthermore, it has been hypothesized that begging is coadapted and potentially genetically correlated with parental care traits as a result of social selection. Further experiments on the causal processes that shape the evolution of begging are therefore essential. We applied bidirectional artificial selection on begging behaviour, using canaries (Serinus canaria) as a model species. We measured the response to selection, the consequences for offspring development, changes in parental care traits, here the rate of parental provisioning, as well as the effects on reproductive success. After three generations of selection, offspring differed in begging behaviour according to our artificial selection regime: nestlings of the high begging line begged significantly more than nestlings of the low begging line. Intriguingly, begging less benefitted the nestlings, as reflected by on average significantly higher growth rates, and increased reproductive success in terms of a higher number of fledglings in the low selected line. Begging could thus represent an exaggerated trait, possibly because parent-offspring conflict enhanced the selection on begging. We did not find evidence that we co-selected on parental provisioning, which may be due to the lack of power, but may also suggest that the evolution of begging is probably not constrained by a genetic correlation between parental provisioning and offspring begging.

All the methods used for collecting the data are described in details in the manuscipt. 

Individual begging tests were performed five days after hatching following a protocol as described in Estramil et al. (2013). Begging was scored combining measures of nestling posture (1: open beak, 2: open beak and head back, 3: open beak and stretched neck and body, 4: open beak, stretched body and stretched legs) and begging duration by calculating the sum of all begging scores per second per nestling according to Kilner (2001).

We measured parental provisioning for each nest 10 or 11 days after hatching of the first nestling, as described in Estramil et al. (2014a) and Fresneau and Müller (Fresneau and Müller, 2016).