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

recorder and an observer who surveyed birds using a point count method. However, these var- ious studies have delivered inconsistent results: from a higher species richness recorded by an observer [18,32], no statistically significant differences in species richness estimated by both methods [33,34], to a higher species richness detected by a recorder [21]. These disagreements may arise from a few methodological limitations, like differently skilled observers involved in such comparisons, habitat type causing differences in detection probabilities by recorder and observer, or distance categories (or unlimited radius) from which birds are recorded by observers during field surveys [19,23,24,30,35–37]. Therefore, to effectively compare bird bio- diversity estimation by human observers and acoustic recorders, studies conducted by highly skilled observers, in a relatively homogenous habitat are required. By applying various detec- tion distance categories for human observers and comparing them with detections made by recorders we may estimate the detection distance of individual species or groups for recorders and find potential species composition for which recorders and observers work equally well. In this way we make possible comparison of results of our study both with studies applying limited and unlimited distance of detection by human observers, and we show how using fixed or unlimited detection distance by human observer changes interpretation of results compar- ing biodiversity estimation by recorders and observers surveying birds at the same time.

In our study we asked whether acoustic monitoring can be an effective alternative for tradi- tional, human-based approach to bird biodiversity monitoring in relatively homogeneous open habitat–agricultural landscape of meadows. Permanent grasslands and meadows covered 34% of the utilised agricultural area in the EU in 2016 (59 million hectares) [38] and are a key habitat for European birds [39]. The species inhabiting them, together with other farmland birds, are among the group of birds whose populations are rapidly declining [40]. Many European Union countries implement agri-environmental schemes (AES) to protect meadow habitats and their biodiversity [41]. Thus, simple, repeatable, and effective, large-scale monitoring of farmland bird species is needed to understand how birds respond to changes in agricultural management and how the conservation activities being undertaken affect particular species and their groups [42–44]. Here we compared similarities and differences of acoustic and traditional surveys by skilled, experienced observers in estimating biodiversity in Central European meadows over distances � 50 m, � 100 m, and unlimited range. The field methodology followed methods currently used for evaluation of effectiveness of agricultural measures dedicated for bird pro- tection in Poland. We analysed the usefulness of acoustic recorders for biodiversity estimations of various groups of birds: (1) all bird biodiversity, (2) songbird biodiversity, (3) meadow bird biodiversity and (4) farmland bird biodiversity. We also examined which of the most common

bird species are underdetected and which are overdetected by recorders in comparison to human-observers, to show potential, species-specific limitations of acoustic method. The goal of our study was to examine whether data collected by highly skilled ornithologists and acous- tic recorders in meadows give similar estimation of overall bird biodiversity and biodiversity of species important for agricultural landscape—an environment in which large-scale, long- term acoustic monitoring could be used as an alternative approach to human-based bird sur- veys. This could result in establishing a monitoring scheme which enables non-specialists, including landowners, to collect field data which that can be analysed in standardized ways to augment traditional survey approaches. Materials and methods Fieldwork We selected 74 recording sites located in meadows in Poland (see S1 Table for exact locations of each recording site). In each recording site, we visually estimated coverage by main habitat PLOS ONE Acoustic bird biodiversity monitoring in agricultural meadows PLOS ONE | https://doi.org/10.1371/journal.pone.0266557 April 8, 2022 3 / 16