This is the authors' abstract. We don't add key points for this paper.
Birdsong performance traits offer a mechanistically grounded route for developing non-invasive indicators of avian physiological condition at ecologically relevant scales. Sexual selection and honest signalling theory predict that complex, energetically costly song traits should be condition-dependent and thus sensitive to stressors such as parasitism and physiological stress. However, it remains unclear whether these relationships scale beyond individuals or single species to the plot level that would matter for monitoring programs. Here, we explicitly frame birdsong traits as a candidate state indicator and ask whether variation in health metrics across forest plots is reflected in corresponding changes in performance-related acoustic traits. We combined blood-derived health metrics with passive acoustic recordings from 17 forest plots in south-west Germany, focusing on five common songbirds ( Erithacus rubecula , Fringilla coelebs , Parus major , Sylvia atricapilla , Turdus merula ). For each species-by-plot combination, we quantified median haemosporidian infection intensity and heterophil-to-lymphocyte ratio. We extracted performance-related song traits using automated acoustic metrics (peak frequency, spectral bandwidth, number of spectral peaks, chroma variance, tonnetz variance) from species-specific, BirdNET-based classifications. Using mixed-effects models at the species-by-plot scale, we tested whether variation in these health metrics correlates with acoustic feature expression. Across species, multiple acoustic traits responded to variation in infection intensity and immune activation, although the direction and magnitude of responses differed among taxa. Importantly, models including health metrics consistently outperformed species-only models, demonstrating that aggregated acoustic trait expression tracks aggregated physiological condition at the plot scale. Although response directions differed among species, models including health metrics consistently outperformed species-only models, demonstrating that health-acoustic relationships are systematic rather than random. These results suggest that acoustic indicators of avian health must be interpreted relative to species-specific baselines rather than assuming a universal response across taxa. Building on this evidence, we propose that performance-related song features serve as candidate indicators of avian physiological condition. Their non-invasive collection via passive acoustic monitoring and sensitivity to health variation make them promising components of scalable monitoring frameworks. Rather than presenting a finalised composite index, this study establishes the conceptual and analytical groundwork for developing health-linked acoustic indicators that complement traditional biodiversity metrics by adding functional information on organismal condition.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Arend et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: