Supplementary feeding and young extraction from the wild are not a sensible alternative to captive breeding for reintroducing bearded vultures Gypaetus barbatus.

1. Reintroduction programmes need regular assessments of their achievements and cost-efficiency so as to adapt management strategy. Ferrer et al. argue that the reintroduction of the bearded vulture, which has so far relied upon the release of captive-reared chicks, could instead use the extra number of young produced by supplementary feeding on poor-quality territories in the Pyrenees. They argue that this strategy would not affect the donor population while being much cheaper than a captive breeding programme. We question this assertion based on several grounds.
2. Ferrer et al. used incorrect data when estimating the effects of supplementary feeding on the production of young. Our reanalyses using the same data source but corrected for missing values show effect sizes of low magnitude and with substantial uncertainty, questioning any positive effect of supplementary feeding on productivity. Moreover, Ferrer et al.’s experiment actually lacks a genuine statistical control due to weak temporal stability in territory quality: we demonstrate that average productivity of territories of low-quality during the baseline period (2001–2006) might actually have been increasing during the ‘treatment’ period (2007–2010) even without food supplementation due to a possible regression to the mean effect.
3. Our demographic reanalysis furthermore suggests that the donor population would not be that resilient to the extraction of young, resulting in trajectories far less optimistic (53–56 pairs in a time horizon of 50 years) than the 70 pairs reported. We also suggest that the costs of a captive breeding programme are substantially lower than suggested by Ferrer et al.
4. Synthesis and applications. The management recommendations proposed by Ferrer et al. appear unjustified. We thus urge governmental agencies to avoid making strategic political decisions and deploying conservation action on such a flawed base. We do not question here the utility of experimental approaches in reintroduction biology, but advocate independent assessment of study designs, data handling and quantitative analyses, notably when extraction of individuals from endangered populations is proposed as the best alternative option.