In this study, we found significant differences in the recorded number of white stork fledglings between the human observer and the UAV. It is not surprising as the access to a nest for humans standing on the ground is much more constrained than for a flying object. However, it is important information that during the long-term censusing of the population, the overall breeding output may be underestimated when only ground checking is performed. In practice, it means that when the observer is counting fledglings standing on the nest, some of them can sit or lay in the nest, invisible to the observer, particularly when the nest is large construction (Vergara et al. 2010; Zbyryt et al. 2021). Moreover, the way of ground survey predicts the probability of making a mistake in recording fledgling numbers, i.e. in colonies, the probability of a mistake was significantly lower than in solitary nests. It is probably due to the different observation angles in these two populations. Solitary nests in Poland are mainly located in villages on electric poles or roofs (Tobolka et al. 2013), so it is sometimes difficult to observe them from a further distance, outside the village, due to rural settlements surrounding the nest constraining the observation. Colonies in Spain are often on tree aggregations far from the settlements. Due to the timid behaviour of nestlings storks, the best solution to observe them is from a distance, which in turn makes the angle of observation more convenient to record the exact number of fledglings, including those sitting or lying in the nest.
On the other hand, when performing long-term monitoring, an additional interview with property holders where the nest is located is a standard action, particularly when brood fails (Janiszewski et al. 2013; Tobolka et al. 2015). Thus, the obtained results can be supplemented and more accurate. What is more, in the case of the white stork, often the monitoring is accompanied by chick banding, i.e. direct visits in the nests. Therefore, data collected on the long-term study sites can be considered relevant for population monitoring. However, the use of UAV monitoring during the time when stork nestlings are hatching would bring much more information when studying breeding biology of the species due to collecting data that human observer cannot record standing on the ground.
The time needed for obtaining the information about the number of fledglings differed significantly between the human observer and the UAV, which contradicts the study on oystercatcher Haemantopus ostralegus, where the traditional method of censusing was significantly more time-consuming and, therefore, more costly than UAV. However, the white stork nests close to human settlements, on buildings, and in prominent locations (Tobolka et al. 2013), which facilitates monitoring compared to ground-nesting birds. Hence, in the case of the white stork, using a UAV does not shorten the time needed for a survey and does not necessarily reduce fieldwork costs. Using a UAV in a constrained period (i.e. two weeks for nestling monitoring) in large areas requires several UAVs and the associated qualified personnel. It requires additional permits when flying close to settlements and electricity networks or close to airports or military training areas. Moreover, the use of UAV is often restricted by current laws, with the impossibility of using this system in particular parts of the study area.
Using the data from Spain, where white storks breed mainly in colonies, the recorded colony size was significantly larger when surveyed by the human observer than by UAV. It contrasts with the study on the ground-nesting bird, the black-headed gull Chroicocephalus ridibundus, where images obtained from an unmanned aircraft system (UAS) allowed to establish colony size with very high precision (Sarda-Palomera et al. 2012) and results of Oystercatcher studies where a higher number of occupied nests was detected using UAS than by traditional census method (Valle and Scarton 2019). In the case of the white stork, the differences in the assessment were due to the tree canopy coverage that affected the numbers obtained by UAV, which is not a problem during a survey in open space areas such as bar ground, beach or sparsely vegetated habitat. However, using a thermal infrared camera would solve this issue as it was proven an effective tool in at least mammalian studies (Linchant et al. 2015).