基于音节聚类分析的被动声学监测技术及其在鸟类监测中的应用

被动声学监测通过分析鸟鸣声信息来实现物种识别,为鸟类多样性监测提供了一种切实可行的技术方案。由于鸟种的鸣声复杂多变,如何通过声纹快速准确辨别物种,分析鸟类丰度,降低对人工操作的需求等技术难题,成为基于声纹的鸟类多样性监测所面临的挑战。本文提出了基于音节聚类的鸟类鸣声监测框架:首先通过音高、频率平坦度等音频特征在声纹数据中提取音节,然后通过无监督表征学习与狄利克雷过程(Dirichlet process)混合模型对音节进行深度无监督聚类训练,完成音节聚类和自动音节种类推断。分析结果表明,本文提出的基于音节聚类的鸟类鸣声监测框架在处理开源数据集白腰文鸟(Lonchura striata)的曲目时可获得接近90%的聚类准确率。在此基础上,本研究对2022年4-5月在广州市白云山公园固定监测点所录制的10种鸟类鸣声进行了无监督的音节聚类分析,验证了本文所提出的基于音节聚类的鸟类鸣声监测框架的有效性:本技术不仅可以支持快速鸟类物种识别,还可以统计和分析不同物种鸟鸣在时间、频度、数量上的变化。这些结果表明,基于音节聚类的鸟类鸣声监测框架可以显著降低对人工标注训练数据的要求,克服传统鸟鸣物种识别框架...     

基于声音指数的鸟类多样性监测

鸟类运动能力强,且对生境变化敏感,常作为环境指示物种。鸣声是鸟类重要的通讯方式,其特征具有物种特异性。随着野外录音及鸣声分析技术和设备的快速发展,从录音中自动提取反映鸣声特征的参数,快速评估鸟种多样性已成为可能。相比于传统的鸟类调查方法,通过录音监测鸟类,具有方便快捷的优点。对鸣声反映鸟类多样性的理论基础、常用的声音指数,以及该领域的关键技术加以介绍,为鸟类监测提供方法借鉴。     

鸟类鸣声初探

对鸟类鸣声的研究是生物声学研究中最为活跃的领域之一,对鸟鸣声的应用在很多方面具有十分广阔的前景。鸟类的鸣声可分为鸣唱和鸣叫,不同鸟类具有不同的鸣叫能力,可据此将鸟类分为鸣禽类和非鸣禽类两大类。文章初步综述了鸟类鸣声的相关知识,对鸣禽和非鸣禽发声器官的结构功能进行了比较。     

基于深度特征融合的鸟鸣识别方法及其可解释性分析

鸟鸣识别是生态监测的重要手段,为进一步提升鸟鸣识别的准确性和鲁棒性,本文提出了1种新的基于深度特征融合的鸟鸣识别方法。该方法首先利用深度特征提取网络对鸟鸣的对数梅尔谱图和补充特征集的深度特征进行提取,再将两种深度特征进行融合,最后使用轻量级梯度提升机(light gradient boosting machine,light GBM)分类器进行分类。本文充分利用深度神经网络的特征提取能力以及light GBM的分类性能,将特征提取和特征分类过程进行分离,从而实现了高准确率的鸟鸣识别。实验结果显示,本文提出的方法在北京百鸟数据集中取得了目前已知的最佳结果,模型的平均准确率达到了98.70%,平均F₁分数达到了98.84%。相比传统方法,深度融合特征在鸟鸣识别任务上准确率提升了5.62%以上。同时,引入的light GBM分类器使分类准确率提升了3.02%。此外,在CLO-43SD和Bird CLEF2022比赛的数据集中,本文方法也展现出卓越的性能,分别取得了98.32%和91.12%的平均准确率。本文还引入了类激活图对不同类型鸟鸣的识别结果进行可解释性分析,揭示了...     

体征对灰眉岩鹀鸣唱特征的影响

声信号是鸟类交流的重要方式。鸟类的鸣声特征受到生境的塑造和性选择,也受到发声器官的影响,如种间的比较研究发现,体重大的鸟类鸣声频率低,喙短的鸟类鸣声频宽大。在种内是否也存在类似的规律,目前的结论还不一致。2012年6月,在北京小龙门地区(40°00'N,115°26'E)测量了17只灰眉岩鹀(Emberiza godlewskii)的形态参数以及鸣声的频率特征,未发现体重或喙长与鸣声的最高频率、最低频率、峰频或频宽存在显著的相关性。本研究的结果支持在小型雀形目鸟类中,体型对繁殖期鸣唱频率特征的影响不显著。     

回放技术在鸟类学研究中的应用

鸟类鸣声复杂,在鸟类的生活史中具有重要意义.自 19 世纪 50 年代末期基于声音回放技术开展鸟类鸣声、行为及生理等研究以来,鸟类学家取得了许多单靠鸣声声谱分析所无法取得的成果.但在中国,回放技术在鸟类学的应用尚处于起步阶段.根据现有的研究资料总结了相对于其他动物而言,回放技术在鸟类学研究中应用更为广泛的原因;对回放技术在种间关系、种内识别、种群数量调查、行为及生理等领域的应用及其意义进行了综述;对回放实验中存在的不稳定因素、假重复问题以及回放反应评估体系进行了讨论;并对回放技术在鸟类分类学及濒危鸟类监测中新的应用以及互动回放技术的发展前景进行了展望.     

鸟类鸣声研究

介绍了鸟类呜声采集和分析的方法,呜声的多样性及其形成原因以及鸟类鸣声在个体识别、地理变异、系统分类等生态生物学研究中的应用。     

环境噪音对鸟类鸣声的影响及鸟类的适应对策

在高噪音环境中生存的动物,发出的声信号会与噪声的频率、振幅和时间等重叠,使动物声信号的传播效率降低.鸟类主要靠鸣声通讯,鸣声传播效率下降会影响鸟类个体间识别、配偶关系、领域防卫、种群密度、群落结构等.本文综述了城市噪音、自然噪音等环境噪声对鸟类鸣声的影响以及鸟类的适应对策,提出在城市化进程中要关注噪音对鸟类的影响,并展望了本领域今后可能的研究热点.     

牡丹鹦鹉(Psittacula agapornis)的鸣声发育

牡丹鹦鹉的鸣声发育分为5个时期。0￣12d为先天性鸣声期,鸣声特征为以基本音(BS)为主音的单音节单音调声,声长短。13￣30d为空白模板形成期,鸣声特征为以基本音为主音的多音节单音调声,声长和音量显著性增加,表明发声学习通路开始形成。31￣45d为鸣声模板形成期,鸣声特征为以第1陪音(UP)1为主音的多音节单变调声,声长和音量显著性增加,主音频提升约920音分,涵盖的律音数增加1倍,表明发声学习模板逐渐形成,发声学习开始。46￣90 d为鸣声反馈学习期,鸣声特征为以BS和UP1为主音的双变调声,声长和音量显著性增加,主音频涵盖的律音数增加2.6￣3.0倍,第2主音频提升约970音分,表明稳定模板逐渐形成,短期记忆逐渐向长期记忆转化。91d后为完美鸣声期,鸣声特征为以UP1为主音的复合变音调声,声长、音量、主音频的提升和涵盖的律音数都趋于平稳,鸣声稳定而和谐,表明发声运动通路基本形成。这些结果可为进一步揭示鸟类发声学习记忆机制提供直接的声学证据。     

大鹰鹃鸣声的日节律

鸟类的鸣唱具有吸引配偶和保卫领域的功能,多为雀形目雄性鸟类发出,在其婚配、繁殖中起重要作用。非雀形目鸟类缺乏内鸣肌,发出的声音较为单调。但杜鹃科(Cuculidae)等少数非雀形目鸟类,利用鸣叫来吸引异性和宣示领域,功能上类似雀形目鸟类的鸣唱。鸟类在繁殖期面临觅食等基本生理活动与求偶行为的时间权衡。而鸣唱是雄性鸟类在繁殖期耗能较多的求偶行为,鸣唱的活跃程度受到外界环境和鸟类自身习性的影响。通过研究鸟类鸣声的日节律,有助于了解鸟类对生活史策略的响应。本研究于2016年和2017年鸟类繁殖季在北京小龙门森林公园(40°00′N,115°26′E)进行。2016年和2017年在大鹰鹃(Hierococcyx sparverioides)活动区利用录音机(美国Wildlife Acoustics公司,型号SM4)分别录制了3 d和43 d的录音。通过Kaleidoscope Pro 4.0.3软件(美国Wildlife Acoustics公司),量化录音并提取反应大鹰鹃鸣声特征的参数,进而自动识别出录音中大鹰鹃的鸣声。在优化识别条件后,对大鹰鹃鸣声识别的正确率可以达到60.26%,探测率可以达到44.71%。发现大鹰鹃有两个鸣叫的高峰,一个在3:00时,另一个在19:00时。与同域分布的其他鸟类相比,大鹰鹃鸣声的高峰时段持续的时间更长,且具有夜间鸣叫的特点。结合大鹰鹃的生活史,我们对其鸣声日节律进行了讨论。     

Singen mit Kopf und Kragen

Singing with the wings – instrumental sound production in birds Bird wing and tail feathers are able to produce sound in flight and during related movements. Noisy sound is produced primarily as an epiphenomenon to the movements. The sound can, however, also include tonal or harmonic structures and can gain signal value when used in social situations. In a similar way, birds can produce sounds by clattering their keratin covered beaks. For the first time, in this article, the acoustic properties of instrumental bird sounds are analyzed using sonograms. Up to now the biological meaning of instrumental sounds in birds is inferred mainly out of the situations observed.     

If a bird flies in the forest,does an insect hear it?

Birds are major predators of many eared insects including moths, butterflies, crickets and cicadas. We provide evidence supporting the hypothesis that insect ears can function as ‘bird detectors’. First, we show that birds produce flight sounds while foraging. Eastern phoebes (Sayornis phoebe) and chickadees (Poecile atricapillus) generate broadband sounds composed of distinct repetitive elements (approx. 18 and 20 Hz, respectively) that correspond to cyclic wing beating. We estimate that insects can detect an approaching bird from distances of at least 2.5 m, based on insect hearing thresholds and sound level measurements of bird flight. Second, we show that insects with both high and low frequency hearing can hear bird flight sounds. Auditory nerve cells of noctuid moths (Trichoplusia ni) and nymphalid butterflies (Morpho peleides) responded in a bursting pattern to playbacks of an attacking bird. This is the first study to demonstrate that foraging birds generate flight sound cues that are detectable by eared insects. Whether insects exploit these sound cues, and alternatively, if birds have evolved sound-reducing foraging tactics to render them acoustically ‘cryptic’ to their prey, are tantalizing questions worthy of further investigation.     

Extracting bird migration information from C‐band Doppler weather radars

Although radar has been used in studies of bird migration for 60 years, there is still no network in Europe for comprehensive monitoring of bird migration. Europe has a dense network of military air surveillance radars but most systems are not directly suitable for reliable bird monitoring. Since the early 1990s, Doppler radars and wind profilers have been introduced in meteorology to measure wind. These wind measurements are known to be contaminated with insect and bird echoes. The aim of the present research is to assess how bird migration information can be deduced from meteorological Doppler radar output. We compare the observations on migrating birds using a dedicated X‐band bird radar with those using a C‐band Doppler weather radar. The observations were collected in the Netherlands, from 1 March to 22 May 2003. In this period, the bird radar showed that densities of more than one bird per km³ are present in 20% of all measurements. Among these measurements, the weather radar correctly recognized 86% of the cases when birds were present; in 38% of the cases with no birds detected by the bird radar, the weather radar claimed bird presence (false positive). The comparison showed that in this study reliable altitudinal density profiles of birds cannot be obtained from the weather radar. However, when integrated over altitude, weather radar reflectivity is correlated with bird radar density. Moreover, bird flight speeds from both radars show good agreement in 78% of cases, and flight direction in 73% of cases. The usefulness of the existing network of weather radars for deducing information on bird migration offers a great opportunity for a European‐wide monitoring network of bird migration.     

Method for passive acoustic monitoring of bird communities using UMAP and a deep neural network

An effective practice for monitoring bird communities is the recognition and identification of their acoustic signals, whether simple, complex, fixed or variable. A method for the passive monitoring of diversity, activity and acoustic phenology of structural species of a bird community in an annual cycle is presented. The method includes the semi-automatic elaboration of a dataset of 22 vocal and instrumental forms of 16 species. To analyze bioacoustic richness, the UMAP algorithm was run on two parallel feature extraction channels. A convolutional neural network was trained using STFT-Mel spectrograms to perform the task of automatic identification of bird species. The predictive performance was evaluated by obtaining a minimum average precision of 0.79, a maximum equal to 1.0 and a mAP equal to 0.97. The model was applied to a huge set of passive recordings made in a network of urban wetlands for one year. The acoustic activity results were synchronized with climatological temperature data and sunlight hours. The results confirm that the proposed method allows for monitoring a taxonomically diverse group of birds that nourish the annual soundscape of an ecosystem, as well as detecting the presence of cryptic species that often go unnoticed.     

虎皮鹦鹉声行为的研究

本文对虎皮鹦鹉(Melopsittacus undelafus)不同声行为叫声的声学特性进行了定量分析,为地震前鸟类声行为的定量观测和鸟声学研究提供了新的资料和认识。 虎皮鹦鹉求偶、离群、受伤及其同伴叫声中变音调声段的声学特性有明显的差异。但是抗议叫声具有鸟类警戒叫声的一般特征,即都有时间图样基本相同的音节组成,每个音节含有若干个调幅脉冲列,其频带和声强都明显加宽和增大。 雏鸟不同的声行为叫声都是简单结构。与成鸟叫声相比较,尤其是抗议叫声呈现明显的发育过程。     

The role of farm structure on bird assemblages around a Kenyan tropical rainforest

Although it is clear that the farmlands neighbouring fragmented forests are utilized by some forest birds, it is not clear how birds in general respond to farmland habitat mosaic. An effort was made to determine how bird density and foraging assemblages were influenced by farm structural characteristics and distance from forest edge. Thirty farms up to a distance of 12 km around Kakamega forest in western Kenya were studied. Farm structure entailed size, hedge volume, habitat heterogeneity, woody plant density, plant diversity and crop cover. Birds were surveyed using line transects and DISTANCE analyses and classified into six feeding guilds and three habitat associations. Size of farms increased away from the forest, as woody plant density, plant diversity, indigenous trees and subsistence crop cover declined. The most important farm structure variable was hedge volume, which enhanced bird species richness, richness of shrub‐land bird species and insectivorous bird density (R = 0.58, P < 0.01). Bird density increased with tree density while indigenous trees were suitable for insectivores and nectarivores. There were very few forest bird encounters. Agricultural practices incorporating maintenance of hedges and sound selection of agroforestry trees can enhance conservation of birds on farmland, though, not significantly for forest species.     

Avian Assemblages at Bird Baths: A Comparison of Urban and Rural Bird Baths in Australia

Private gardens provide habitat and resources for many birds living in human-dominated landscapes. While wild bird feeding is recognised as one of the most popular forms of human-wildlife interaction, almost nothing is known about the use of bird baths. This citizen science initiative explores avian assemblages at bird baths in private gardens in south-eastern Australia and how this differs with respect to levels of urbanisation and bioregion. Overall, 992 citizen scientists collected data over two, four-week survey periods during winter 2014 and summer 2015 (43% participated in both years). Avian assemblages at urban and rural bird baths differed between bioregions with aggressive nectar-eating species influenced the avian assemblages visiting urban bird baths in South Eastern Queensland, NSW North Coast and Sydney Basin while introduced birds contributed to differences in South Western Slopes, Southern Volcanic Plains and Victorian Midlands. Small honeyeaters and other small native birds occurred less often at urban bird baths compared to rural bird baths. Our results suggest that differences between urban versus rural areas, as well as bioregion, significantly influence the composition of avian assemblages visiting bird baths in private gardens. We also demonstrate that citizen science monitoring of fixed survey sites such as bird baths is a useful tool in understanding large-scale patterns in avian assemblages which requires a vast amount of data to be collected across broad areas.     

The effect of land type diversity and spatial heterogeneity on farmland birds in Norway

The decline in farmland birds observed throughout Europe during recent decades has attracted much attention. Agricultural intensification or land abandonment are commonly forwarded as key drivers. Several countries have established agri-environmental schemes (AES) to counter these negative trends among farmland birds. This paper reports a study of the relationship between land use and bird species in the agricultural landscape of Norway. The main objective was to investigate the effect of spatial heterogeneity and diversity of land use on total richness and abundance of farmland birds at a national level.Monitoring the distribution and abundance of birds is part of the Norwegian monitoring programme for agricultural landscapes. The monitoring programme is based on mapping of 1 × 1 km squares distributed across the entire agricultural landscape. Within these squares permanent observation points are established for bird monitoring. Detailed interpretation of aerial photographs provides the land classification. We tested the relationship between landscape metrics at different levels of land type detail and species richness and abundance of farmland and non-farmland birds.There was a positive relationship between species richness and abundance of farmland birds and agricultural area. For non-farmland birds the relationship was negative. Spatial heterogeneity of land use was a significant positive factor for both farmland and non-farmland species. High land type diversity was positive for farmland bird richness, but negative for abundance. Non-farmland bird richness was not affected by land type diversity, but abundance had a negative response.The results presented in this paper highlight the importance of a spatial heterogeneous landscape. However, we also found that land type diversity could negatively affect the abundance of both farmland and non-farmland birds. Our findings suggest a need for different management approaches depending on whether the aim is increased species richness or abundance. Achieving both aims with the same means might be difficult. We thus suggest a need for land use analyses before proper management strategies can be implemented.     

广西猫儿山自然保护区的兽类和鸟类多样性初步调查-基于红外相机监测数据

红外相机技术是监测兽类和鸟类的重要方法,目前已广泛应用于野生动物资源调查、种群评估和行为生态学等研究领域。2014年7月至2015年1月期间,在广西猫儿山国家级自然保护区布设51个红外相机监测位点（分别位于4种生境和7个海拔区域）,对区内大中型兽类及林下鸟类资源多样性进行了初步调查。研究期间共记录到兽类17种、鸟类30种,包括2种国家I级和7种II级重点保护野生动物。小泡巨鼠（Leopoldamys edwardsi）、红腿长吻松鼠（Dremomys pyrrhomerus）、隐纹松鼠（Tamiops swinhoei）、野猪（Sus scrofa）和赤腹松鼠（Callosciurus erythraeus）的相对丰富度居于兽类前5位,白鹇（Lophura nycthemera）、紫啸鸫（Myophonus caeruleus）、红嘴相思鸟(Leiothrix lutea)的相对丰富度居于鸟类前3位。在不同生境和海拔,兽类和鸟类物种数、多样性指数和均匀度指数存在显著差异。混交林中兽类和鸟类的物种数量、多样性和均匀度指数明显高于其它生境;海拔1400-1600 m鸟类物种数明显高于其它海拔区域。本次监测结果初步了解了猫儿山保护区内大中型兽类及林下鸟兽物种组成、相对数量及空间分布,为后期的科研工作及保护管理提供了丰富的基础资料。