Census methods for White stork (Ciconia ciconia): bias in sampling effort related to the frequency and date of nest visits

Abstract

Periodic censuses have been widely used to identify the population numbers and conservation status of many bird species. In order to be comparable, a census must be homogeneous through time. However, this requirement is not always possible. For this reason, studies addressing a possible bias in sampling efforts are very useful to correct such errors. In the present work, a standardized periodic monitoring of breeding White stork (Ciconia ciconia) at six Spanish colonies was conducted to estimate survey accuracy. We estimated the percentage of breeding pairs and productivity (number of chicks), i.e., accuracy, detected in each possible combination of number of visits as compared with the results obtained for the whole intensive monitoring, i.e., reality. Our results showed that single visits resulted in lower percentages of the number of breeding pairs and productivity detected compared with combinations of two or more visits. Nevertheless, one visit in a single month (April for the number of breeding pairs and June for productivity) did not show significantly lower results than the rest of the combinations of two or more visits. Early or late visits in the season might underestimate breeders by not accounting for either late-occupied or failed nests, respectively. In addition, the obvious increase in the probability of detection related with the number and the size of chicks is probably the reason why later visits in the season reported the highest value of productivity. In conclusion, the estimation bias presented in this study may be used to adjust sampling efforts in the census of the White stork.