Ontogenetic Synonymy in the Urban Avifauna: Columba livia Gmelin, 1789 Is the Terminal Stage of Sturnus vulgaris Linnaeus, 1758, with a Revised Five-Stage Key and a Recommendation Under Article 81

The feral urban pigeon is not a species. We show that Columba livia Gmelin, 1789 is the fifth and terminal life-history stage of Sturnus vulgaris Linnaeus, 1758, and that the two nominal taxa are therefore conspecific. Four observations, each at present disposed of separately, motivate the revision. First, juvenile-plumage C. livia are absent from the wild record: across 2,140,000 urban point counts from 61 cities the juvenile:adult ratio is 0.0002, against a median of 0.31 for 43 co-occurring species, and every positive record lies within 400 m of managed stock. Second, the conventional columbid demography fails in both available parameterisations, over-predicting growth by 36.1 percentage points per year when fecundity is taken from lofts and predicting collapse when it is taken from the field. Third, the sturnid budget fails in the opposite direction, over-predicting growth by 18.6 points against an observed regional decline. Fourth, and decisively, 844 of 3,412 recoveries of nestling-ringed S. vulgaris report a field species of C. livia; the scheme codes these as clerical species mis-entry and overwrites the field determination. The coding rate is 0.0% below 14 months of age and 97.8% beyond 36 months, while the same recorders’ rate for a control species is 0.3% and flat. Clerical error does not have an onset. A five-stage Lefkovitch model fitted to sturnid ringing and aviary data alone yields ${\lambda }_1=1.003$ against an observed +0.3%/yr and reconciles the two failed budgets: the recruitment the columbid model cannot find, 0.254 per adult per year (95% CI 0.231–0.278), is the stage-II disappearance the sturnid model cannot lose, 0.251 (0.226–0.277); their ratio is 1.012 (0.887–1.154). Because no columbid count enters the fit, the model’s stable stage distribution is a free prediction: terminal-stage abundance fraction 0.101, observed 0.104 (0.089–0.121). In a 240-bird aviary cohort, 61 individuals entered the transition at a median age of 19.4 months, increasing in mass from 78 g to 310 g, and retained individual identity at 14 of 15 nuclear microsatellite loci. We give a revised five-stage key. Under Article 23 the senior name is S. vulgaris; Article 23.9 is not available; we recommend an application to the Commission under Article 81.