Based on a press release by the Friedrich Schiller University Jena
Bird communities in forests change significantly over the course of the year. Nevertheless, the differences between various forest types remain clearly discernible across the seasons. This is shown by a new study conducted by researchers from Friedrich Schiller University Jena, the German Centre for Integrative Biodiversity Research (iDiv) and the Northwest German Forest Research Institute. For the study, the research team used passive acoustic recordings and analysed them using artificial intelligence. The results were published in the journal Forest Ecology and Management.
Bird monitoring extended beyond spring
In traditional bird monitoring, ornithologists survey designated areas and record the species they see or hear. Such mapping usually takes place in spring, typically around sunrise, when many species are particularly active in their singing. “However, this method primarily provides insights into bird communities during the breeding season,” says first author Esther Felgentreff, doctoral researcher at the University of Jena and iDiv. “With these methods, forest bird communities at other times of the year are not considered.”
As part of her doctoral thesis, Esther Felgentreff and the research team have now significantly expanded this perspective. In forests around Jena, the researchers employed what is known as passive acoustic monitoring. To this end, they installed small “AudioMoth” recording devices in the study areas, which recorded short sound sequences at regular intervals throughout the year—around the clock, four times a year, for a period of 30 days each time.
AI analyses large volumes of bird songs
The audio recordings were analysed using BirdNET, a freely available, AI-powered method for identifying bird songs. Whilst the application is best known to many as an app, it can also be used for the automated analysis of large datasets. In the Jena study, the automated results were additionally checked manually on a random sample basis: for each species recorded, the research team validated up to 45 detections. In total, 58 bird species were identified in this way.
“It is precisely this combination of automated analysis and expert review that makes the method particularly interesting for research,” emphasizes iDiv member Prof. Dr Markus Bernhardt-Römermann, who is supervising Esther Felgentreff’s doctoral thesis. “This allows bird communities to be monitored over longer periods and on a larger scale with relatively little effort.”
Seasons have a strong influence—yet forest types remain distinguishable
As expected, the seasons proved to be the most significant factor influencing the composition of bird communities. This is primarily linked to the migratory behaviour of many species: numerous breeding birds leave the region in autumn and spend the winter in more southerly areas, whilst only a fraction of the species remains in the area year-round.
Within each season, there was little difference in species numbers between deciduous, mixed deciduous and coniferous forests. However, there were clear differences in species composition. For example, the crested tit and the coal tit are typical of coniferous forests, whilst cavity-breeding species such as woodpeckers, tree creepers and nuthatches are more dependent on deciduous and mixed deciduous forests with deadwood.
The findings thus also highlight the importance of decisions made in forest management for bird diversity, particularly regarding the choice of tree species—especially against the backdrop of the current forest restructuring. A mosaic of different forest structures could help to promote high bird diversity throughout the year.
Promising for research and nature conservation
According to the researchers, the study shows that passive acoustic monitoring holds great potential for biodiversity research. The cost-benefit ratio is particularly favourable: species occurrence can be recorded in a standardized manner, repeatedly and over long periods of time.
Particularly at a time when biodiversity is changing, and knowledge of species is dwindling, such methods could help to expand monitoring programmes. At the same time, specialist knowledge remains indispensable—for instance, for validating the data and interpreting its ecological significance. In the future, this approach could also be applied to other vociferous animal groups, such as bats or grasshoppers.
The study was part of the long-term monitoring project ‘WZE@Jena’, which involves conducting a forest condition survey and various biodiversity monitoring activities across 22 forest sites around Jena. The study was funded by the German Federal Ministry of Research, Technology and Space.
Original publication:
(Researchers with iDiv affiliation and alumni bold)
Felgentreff, E. S., Singer, D. & Bernhardt-Römermann, M. (2026): Forest type consistently shapes bird communities across seasons: Insights from passive acoustic monitoring. Forest Ecology and Management 609, 123617. DOI: 10.1016/j.foreco.2026.123617, https://www.sciencedirect.com/science/article/pii/S0378112726001155
Contact:
Esther Felgentreff
Institute of Biodiversity, Ecology and Evolution of Friedrich Schiller University Jena
Senckenberg Institute for Plant Form and Function Jena (SIP)
German Centre for Integrative Biodiversity Research Halle-Jena-Leipzig – iDiv
email: esther.felgentreff@uni-jena.de
Prof. Dr Markus Bernhardt-Römermann
Institute of Biodiversity, Ecology and Evolution of Friedrich Schiller University Jena
Senckenberg Institute for Plant Form and Function Jena (SIP)
German Centre for Integrative Biodiversity Research Halle-Jena-Leipzig – iDiv
email: markus.bernhardt@uni-jena.de
Dr. Volker Hahn
Head of Impact and press spokesman
German Centre for Integrative Biodiversity Research Halle-Jena-Leipzig – iDiv
telephone: +49 341 97 33197
email: volker.hahn@idiv.de
Please note: Use of the pictures provided by iDiv is permitted for reports related to this media release only, and under the condition that credit is given to the picture originator.