Subsequent to the publication of the following paper in Ardeola, the author was made aware of an error in the dataset. After re-running the statistical tests the new dataset failed to uphold the previous conclusions and the author has requested that this manuscript should be retracted.
Tellería, J. L. 2014. Has the number of european robins Erithacus rubecula wintering in Spain decreased? Ardeola 61: 389–391. (doi: 10.13157/arla.61.2.2014.389)
In this note I explored whether the number of recoveries of foreign european robins Erithacus rubecula had decreased after controlling for the effect of ringing activity in Spain (a capture per unit effort approach). The results suggested a decline in the number of ring recoveries of long-distance migrant individuals (> 1000 km) and a marked increase in short-distance and local recoveries (< 1000 km). Just after this note was published, it was detected that some dead birds had been included in the dataset. I used a working file contaminated with ring records of birds other than “alive and probably healthy and released by ringer” (condition 8, circumstance 20, in the EURING exchange code).
After analysing the correct records from the last three decades (the best data for exploring these trends given the large number of ringed robins), the results do not support a temporal decrease in ring recoveries of longdistance migrants (table 1). This is the clear result. It may be argued that this result does not agree with the numerical increase in the european robin populations over recent decades, suggesting a reduction in the arrival of migratory individuals. but such a conclusion should be supported by at least two methodological conditions: 1) The ability of ringing effort in Spain to detect changes in the abundance of long-distance migrant robins. This condition seems to be supported in the case of short-distance and local recoveries that track the numerical increase in the Spanish robin population (more ring recoveries than predicted by ringing effort were detected during the past decade; table 1). 2) The number of marked robins mirrors the numerical trends of the species in europe. To test this, it is essential to control for the effect of the number of individuals ringed in europe on the number of ring recoveries. Unfortunately, the number of robins ringed per decade in many countries from which robins come to Spain is not available ( www.euring.org). Obviously, these interpretations and speculations require further investigation and more data.
Decadal changes in short- (< 1000 km) and long-distance (> 1000 km) recoveries of european robins in Spain selected according to condition 8 and circumstance 20 in the EURING exchange code. The results of χ2 analyses to test whether the temporal changes in the number of ring recoveries were predicted by the temporal distribution of ringing effort in Spain are reported. The predicted recaptures (in parentheses) and the number of total birds ringed in Spain per decade are also shown.
[Cambios por década en las recapturas de petirrojos de corta (<1000 km) y larga (>1000 km) distancia seleccionadas de acuerdo con la condición 8 y circunstancia 20 del código de EURING. Se dan los resultados de los análisis de χ2 donde se comprueba si las recapturas observadas difieren de las predichas (en paréntesis) por el esfuerzo de muestreo medido por el número de petirrojos anillados.]
I apologise to readers, reviewers and editors of the journal, and I hope that this letter will serve to prevent the dissemination of some unsound conclusions derived from the note.