Introduction
Many seabird populations face existential challenges as their natural breeding and foraging areas shrink in the face of human activity. BirdLife International (2010) reports that 5% of seabird species are critically endangered, while 28% are threatened. To face this problem the European Commission has put in place the Natura 2000 network. This network consists of “core breeding and resting sites for rare and threatened species” (European Commission, 2019). The monitoring and censusing of seabird species populations in the Natura 2000 areas will help evaluate the initiative’s effect on restoring seabird populations in Europe. More generally, the bird population counts also provide a proxy measure of ecosystem health, as their high mobility allows them to move from disturbed habitat to more attractive locations (Gelinas, 2018).
The task of monitoring and inventorying seabird populations, however, can be difficult as traditional bird censusing methods involve sending an experienced surveyor to count, from the ground, the number of individuals and/or nests in a flock. Such methods are time-consuming, costly, and have been noted to be disruptive to birds (Borrelle & Fletcher, 2017).
There is increasing interest in the field of bird population ecology to leverage new technologies to obtain better, cheaper, and/or more frequent data. Specifically, the use of drones as a means of estimating population counts has been explored by many authors (e.g. Sardà-Palomera et al., 2017, 2012; Chabot, Craik & Bird, 2015; Dulava, Bean & Richmond, 2015; Goebel et al., 2015;Ratcliffe et al., 2015; Rümmler et al., 2015; Chabot & Bird, 2012).
Questions remain as to whether drone surveys cause less disturbance to bird populations than do traditional ground surveys. There is evidence that, for certain applications, drones can have little to no disturbance when compared to ground surveys (e.g. Sardà-Palomera et al., 2017, 2012).
Other studies, however, have shown that the use of drones can cause subjects to become disturbed and vigilant (Rümmler et al., 2015).
The question of whether drones cause significant disturbance to birds is also complicated by the fact that most of the available insight on the question has come from studies on nesting bird colonies (e.g. Chabot, Craik & Bird, 2015; Ratcliffe et al., 2015; Sardà-Palomera et al., 2012). Nesting birds have less of a tendency to flee or to flush because they have a vested interest in protecting their nests and eggs. This fact is important because flushing behaviour is the primary outward signs of disturbance and the one most often used to measure disturbance. This protective behaviour of breeding birds may bias the results of studies attempting to answer the question of whether drones cause significant disturbance to birds.
Intertidal flats are important feeding grounds for migratory and permanent seabird populations. The southern Dutch province of Zeeland has the particularity of having the entirety of its intertidal zones included in the Natura 2000 network. The monitoring and censusing of permanent and migratory seabird population in the Zeeland intertidal flats therefore represent an important step in the assessment of seabird health throughout the Natura 2000 network.
Further, most studies (e.g. Sardà-Palomera et al., 2017, 2012; Chabot, Craik & Bird, 2015; Goebel et al., 2015; Ratcliffe et al., 2015; Rümmler et al., 2015; Chabot & Bird, 2012) in this area have focused on counting individuals in flocks or colonies of a single species. Intertidal flats, however, often host multiple species of seabirds (e.g. sandpipers, oystercatchers, spoonbills). Therefore, when monitoring seabirds feeding on intertidal flats, it is necessary to not only count the birds, but also to identify their species. Dulava, Bean and Richmond (2015) explored the possibility of using drones to distinguish between waterfowl species. They concluded that image resolutions could be no coarser than about 5 mm if the waterfowl species were to be distinguished. Such a resolution requirement for species identification in drone surveys is quite high and would require lower flight altitudes than single-species drone surveys.
The viability of using drones to monitor seabird populations on intertidal flats is then hinged on the question of whether it is possible to obtain imagery of sufficiently fine resolution to identify the species of seabirds at a flight altitude which does not disturb the subjects.