Citizens’ field observations are increasingly stored in accessible databases, which makes it possible to use them in research. Citizen science (CS) complements the field work that must necessarily be carried out to gain an understanding of any of bird species’ ecology. However, CS data holds multiple biases (e.g. presence only data, location error of bird observations, spatial data coverage) that should be paid attention before using the data in scientific research.

The use of Airborne Laser Scanning (ALS) enables investigating forest bird species’ habitat preferences in detail and over large areas. In this study the breeding time habitat preferences of 25 forest bird species were investigated by coupling CS observations together with nine forest structure parameters that were computed using ALS data and field plot measurements. Habitat preferences were derived by comparing surroundings of presence-only observations against the full landscape. Also, in order to account for bird observation location errors, we analysed several buffering alternatives.

The results correspond well with the known ecology of the selected forest bird species. The size of a bird species’ territory as well as some behavioural traits affecting detectability (song volume, mobility etc.) seemed to determine which bird species’ CS data could be analysed with this approach. Especially the habitats of specialised species with small or medium sized territories differed from the whole forest landscape in the light of several forest structure parameters. Further research is needed to tackle issues related to the behaviour of the observers (e.g. birdwatchers’ preference for roads) and characteristics of the observed species (e.g. preference for edge habitats), which may be the reasons for few unexpected results.

Our study shows that coupling CS data with ALS yield meaningful results that can be presented with distribution figures easy to understand and, more importantly, that can cover areas larger than what is normally possible by means of purpose-designed research projects. However, the use of CS data requires an understanding of the process of data collection by volunteers. Some of the biases in the data call for further thinking in terms of how the data is collected and analysed.