The usefulness of unmanned aerial vehicles (UAV) in white stork censusing

Abstract

Unmanned aerial vehicles (UAV) are effective and often noninvasive tools successfully used in bird monitoring. However, when handling long-term data in the context of population changes, the consistency of methods over time is essential as the method-related bias may lead to wrong conclusions. In two distinct populations of white stork Ciconia ciconia, in Poland and Spain, we compared two censusing methods: traditional observation by a human from the ground and using UAV. We recorded the number of recorded fledglings, the time needed to obtain this information, and the number of detected breeding pairs in colonies. We investigated 57 and 117 nests in Poland and Spain, respectively. In Poland and Spain, the number of fledglings was significantly lower when recorded by human observer than by UAV, i.e., 2.21 vs. 2.60 and 1.35 vs. 1.55. The probability of mistakenly recording the number of fledglings by the observer was significantly lower in colonial white storks in Spain than in solitary nesting in Poland. The mean time needed to record the number of fledglings was significantly longer when using a UAV than by a human observer in both populations. The mean number of detected nests in colonies in Spain differed significantly between the human observer and UAV, 13.1 vs. 7.4, respectively. The difference between human and UAV in recorded pairs was higher when colonies were on trees than on human-made structures. We conclude that introducing UAVs in long-term studies may affect the results and should be performed cautiously.