Acoustic approach as an alternative to human-based survey in bird biodiversity monitoring in agricultural meadows

example, implemented to evaluate the effectiveness of agri-environmental schemes undertaken by many European Union countries to protect bird biodiversity in farmland landscapes or any other protection programs which need to measure the effect of undertaken activities. Such citi- zen-science approach could involve landowners or volunteers collecting sound samples using digital recorders (e.g., mobile phones) on their parcels before and during the undertaking of the agri-environmental commitment, and then send recordings to the units responsible for management and monitoring of the effectiveness of agri-environmental measures, responsible for data analysis. Applying acoustic monitoring of agriculture landscape is possible through different scenarios. All of them need planning and management at regional or national level, in accordance with current guidance for acoustic monitoring of environment [9–12]. In this study we showed that using acoustic monitoring we can estimate farmland and meadows bird biodiversity just as well as when using traditional human-based approach. The advantage of acoustic monitoring is that data are collected and analysed in standard- ised way, the costs of acoustic monitoring are lower in comparison to traditional human-based approach [10], and it is possible to increase survey duration to get more accurate and precise estimation of population parameters [23]. Conclusions Our study showed that: • Highly skilled human observers surveying birds within unlimited radius detect more species

than acoustic recorders, because acoustic detection range of observers is longer than record- ers and observers may detect visually species which are silent. • Meadows and farmland bird biodiversity is equally estimated by recorders and human observers surveying birds within 100 m radius. Therefore, it is possible to replace traditional, human based point-counts by acoustic approach without losing effectiveness in detection of the species which are the most common and typical for agriculture landscape. • Although recorders do not detect some silent and very distant species, acoustic approach can be effectively adapted to long-term, large-scale monitoring of birds in agriculture landscape. Supporting information S1 Fig. Detection rate (± SE) of 48 the most common bird species by four detection meth- ods: (1) recorder, (2) observer up to 50 m, (3) observer up to 100 m and (4) observer in unlimited distance. The number of surveys (N) at which the species was detected indepen- dently on the method of detection is given. Results of Wilcoxon two-related simple test are given. Tests compare differences in detection rate by recorder and observers surveying birds within 50 m, 100 m and unlimited distance. (PDF) S1 Table. Summary dataset, containing data from a field survey (soundscape recording and point-counting). Observer, recording site ID, geographical coordinates, survey (early– from April 11 to May 20, hours: 05:08–09:53 AM; late–from May 27 to June 30; hours 05:01– 08:59 AM, local time), detection method, number of detected all bird species, songbird species,

meadow bird species and farmland bird species, visibility around the recording point are given. (XLSX) PLOS ONE Acoustic bird biodiversity monitoring in agricultural meadows PLOS ONE | https://doi.org/10.1371/journal.pone.0266557 April 8, 2022 12 / 16