Data from: Repertoire-based individual acoustic monitoring of a migratory passerine bird with complex song as an efficient tool for tracking territorial dynamics and annual return rates

In field ecological and behavioural studies, it is often necessary to identify specific individuals. In birds, colour rings are frequently used to mark individuals; however, rings are often difficult to observe, especially in small species and dense habitats. Acoustic-based monitoring detecting individuals by their characteristic vocalization is a potentially suitable alternative, but this approach is challenging in species with complex songs. On the example of the Tree Pipit (Anthus trivialis), a small migratory passerine often singing in flight or from perches obscured by foliage, we demonstrate that acoustic monitoring based on the syllable repertoire can be very efficient tool for individual recognition. During a 3-year study, we obtained over 500 recordings from males from one study population (a number of them returning after winter). Males banded with colour rings were repeatedly recorded throughout the seasons, and syllable repertoires were determined from spectrograms for each recording. The repertoire of each unambiguously identified male was distinct and stable within as well as between seasons; and males with similar syllable repertoires differed in syntax. Based on the congruence between identification based solely on spectrogram assessment, and that based on observation of colour rings, we inferred that reliable identification of singing males (including non-ringed ones) was possible in the studied population from assessing a repertoire and song syntax of <5-min recording (containing 20–30 songs). The acoustic-based data: (i) increased the overall estimated number of territorial males at the study locality (from 49 ringed to 61), and improved the estimates of the period of their presence; (ii) revealed dynamic within-season changes in territory occupancy that would otherwise be missed; and (iii) allowed identification of returning birds (including non-ringed ones and those actively avoiding approaching humans). Our results suggest that some commonly used methods may substantially underestimate return rates of migratory bird species. Individual acoustic monitoring should be applicable on various bird species with complex song and stable repertoires, and may be particularly useful for those living in dense habitat or sensitive to handling.

在野外生态与行为学研究中，通常需要对特定个体进行识别。在鸟类研究中，研究者常使用彩色环志（colour rings）对个体进行标记；但环志往往难以被观测到，尤其是针对小型鸟类以及栖息于茂密生境中的类群。基于声学的个体识别监测技术，即通过物种特有的鸣唱进行个体甄别，是颇具潜力的替代方案，但对于鸣唱结构复杂的鸟类类群而言，该方法仍存在较大挑战。
以林鹨（Anthus trivialis）——一种常于飞行中或被枝叶遮挡的栖木上鸣唱的小型迁徙雀形目鸟类——为研究对象，本研究证实基于鸣唱音节曲目（syllable repertoire）的声学监测可成为高效的个体识别工具。本研究为期三年，从单一研究种群的雄性个体中获取了超过500条鸣唱录音，其中部分个体为越冬后返回的留驻个体。佩戴彩色环志的雄性个体在多个季节中被多次录音，我们通过每条录音的语谱图（spectrograms）解析得到其鸣唱音节曲目。
经明确识别的每只雄性个体，其鸣唱音节曲目在单季内与跨季间均表现出显著特异性与稳定性；即便鸣唱音节曲目相似的雄性个体，其鸣唱句法（syntax）也存在差异。基于仅通过语谱图评估得到的个体识别结果，与通过彩色环志观测得到的识别结果之间的一致性，我们推断：在本研究种群中，仅需对时长不足5分钟、包含20~30段鸣唱的录音进行曲目与鸣唱句法分析，即可实现对鸣唱雄性个体（包括未佩戴环志的个体）的可靠识别。
声学监测获取的数据具有以下优势：(1) 提升了研究区域内领域雄性个体的总估算数量（从49只佩戴环志的个体增至61只），并优化了其驻留时长的估算结果；(2) 揭示了季内领域占据情况的动态变化，而这类变化若采用传统方法则会被遗漏；(3) 可实现对越冬返回个体的识别，包括未佩戴环志的个体以及主动躲避人类靠近的个体。
本研究结果表明，部分常用的鸟类个体识别方法可能大幅低估了迁徙鸟类的越冬返回率。基于个体声学监测的识别技术，可推广应用于各类鸣唱结构复杂且音节曲目稳定的鸟类类群，对于栖息于茂密生境或对捕捉操作较为敏感的物种而言，该方法尤其具有应用价值。