| 样点法 Point counts | 声学监测 Acoustic monitoring | 参考文献 Reference | | 优势Advantage | (1)最常采用的鸟类调查方法之一, 易于实施、易于做到随机化或系统化。One of the most commonly used bird survey methods, easy to implement and easy to randomize or systematize.
(2)观测过程中能够直接获得鸟的种类、数量、行为活动等信息, 具有高效性和灵活性。The observation process is efficient and flexible as it provides direct access to information on bird species, numbers, behavioral activities, etc.
(3)适用于复杂、斑块化的生境。相较于样线法, 观察员有更多的时间观察鸟类, 减弱行走速度带来的影响。Suitable for complex habitats, observers have more time to observe birds and attenuate the effects of walking speed compared to the sample line method. | (1)快速发展的调查方法之一。操作简单易重复, 支持同时进行大尺度、长期、连续的动态监测与跟踪。One of the rapidly developing survey methods, simple and repeatable in operation, supports simultaneous large-scale, long-term, continuous dynamic monitoring.
(2)不侵入自然环境, 尽可能地减少现场调查对动植物的影响。有利于对偏远地区、稀有物种的监测。It does not invade the natural environment, minimizes the impact of field surveys on flora and fauna, and facilitates the monitoring of remote areas and rare species.
(3)录音永久保存、可反复监听, 降低识别偏差。声学指数和机器学习的不断发展有助于快速评估生物多样性的动态变化。Recordings can be permanently stored and repeatedly monitored to reduce identification bias. The development of acoustic indices and machine learning helps to rapidly assess dynamic changes in biodiversity. | 吴飞和杨晓君, 2008; Zhang et al, 2018; Darras et al, 2019 | | 劣势Weakness | (1)观测次数有限, 不能保证多样点间同步进行, 观测周期影响结果的可靠性。The limited number of observations does not guarantee the synchronization of multiple points, and the period affects the reliability of the results.
(2)对观察员的专业能力要求较高。在地形复杂、植被密集的森林中, 因视线容易受阻导致记录不完全。The professional competence of observers is required. In forests with complex terrain and dense vegetation, it is easy to have incomplete records due to obstructed vision.
(3)人类的接近引入了惊飞或回避效应, 尤其是多名调查人员带来的干扰可能会使鸟类远离观察者的视线, 影响调查结果。Human proximity introduces a startle or avoidance effect, especially as disturbance from multiple investigators may keep birds away and affect findings. | (1)应对大数据的挑战。长期录制产生庞大的音频文件, 需要考虑大数据的存储与处理, 电池容量及更换成本等。The challenge of big data. Long-term recording generates huge audio files, requiring consideration of big data storage and processing, battery capacity, and replacement costs, etc.
(2)人工识别需要大量的时间精力, 目前自动识别的技术还易受到背景噪声和声音重叠的干扰导致结果偏差。 Manual recognition takes a lot of time, and current automatic recognition techniques are still subject to interference from background noise and sound overlap, leading to biased results.
(3)天气条件或人为破坏对录音设备的灵敏度造成影响, 长期监测中数据丢失可能会在很长一段时间内不被注意到, 需要定期查看维护。Weather conditions or vandalism affect the sensitivity of recording equipment, and data loss during long-term monitoring may unnoticed and require regular maintenance. | Prabowo et al, 2016; Priyadarshani et al, 2018; Darras et al, 2019 |
|