生物多样性

• • 上一篇    下一篇

北京中心城区公园中生物声对植被三维结构的响应

白梓彤1,2, 王成1,2*, 齐志勇3   

  1. 1.中国林业科学研究院林业研究所, 北京 100091; 2.国家林业和草原局城市森林研究中心, 北京 100091; 3.北京大学地球与空间科学学院遥感与地理信息系统研究所, 北京 100871
  • 收稿日期:2025-06-10 修回日期:2025-08-18 接受日期:2025-11-29
  • 通讯作者: 王成
  • 基金资助:
    中央级公益性科研院所基本科研业务费专项资金经费项目(CAFYBB2020ZB008); 国家重点基金项目(2021YFE0193200)

Biophony responses to different vegetation structure in urban central parks of Beijing

Zitong Bai, Cheng Wang, Zhiyong Qi   

  1. 1 Research Institute of Forestry, Chinese Academy of Forestry, Beijing 100091, China 

    2 Urban Forestry Research Center, National Forestry and Grassland Administration, Beijing 100091, China 

    3 Institute of Remote Sensing and Geographic Information System, School of Earth and Space Sciences, Peking University, Beijing 100871, China

  • Received:2025-06-10 Revised:2025-08-18 Accepted:2025-11-29
  • Contact: Cheng Wang
  • Supported by:
    National Non-Profit Research Institutions of the Chinese Academy of Forestry(CAFYBB2020ZB008); National Key R&D Program of China(2021YFE0193200)

摘要: 植被三维结构显著影响公园绿地中声景的形成与空间分布, 生物声作为声景的重要组成部分, 能够间接反映区域的生物多样性水平及生态系统健康状况。然而, 目前关于城市公园生物声对植被三维结构特征的响应机制仍缺乏系统的研究。本研究于2024年夏季在北京城市中心公园的52个样点中同步采集了声音数据和背包激光雷达植被数据。基于现场录制的声音计算了6个典型声学指数, 并从激光雷达点云中提取了42个植被结构变量。利用主成分分析和XGBoost-SHAP模型, 明确对生物声产生显著影响的关键植被结构变量及其重要性, 并使用广义加性模型(generalized additive model, GAM)量化边际贡献与生物声特征之间的非线性响应关系。主要结果如下: (1)不同生物声频段的功率谱密度(power spectral density, PSD)夏季的日变化模式存在差异, PSD2–4 kHz与PSD4–6 kHz频段表现出相似的日节律特征, 而PSD6–10 kHz则呈现出错峰鸣唱的时间分布。(2)平均胸径、冠层起伏比和粗糙度指数是驱动各频段生物声表达的重要植被结构因子; 而林下植被结构和冠层覆盖度则对声学复杂度指数(acoustic complexity index, ACI)、声学多样性指数(acoustic diversity index, ADI)和生物声多样性指数(bioacoustic index, BIO)的调控起到关键作用。(3)适度疏透的林分结构与中等尺度的林木大小更有利于鸟鸣声的增加, 冠层表面形态对昆虫声具有显著调控作用, 冠层起伏比的增加普遍促进生物声的表达。(4)林下植被结构过密, 寡光区体积过大不利于多声源的共存与传播, 会削弱声景的多样性。本研究系统地揭示了植被三维结构对生物声的复杂影响机制, 明确了在生物声营造中的关键植被结构因子, 为绿地中的声景优化和生物多样性保护提供了科学依据。

关键词: 声景监测, 声学指数, 激光雷达, 城市森林, 公园

Abstract

Aims: The three-dimensional structure of vegetation is crucial for the formation and spatial distribution of soundscapes in green spaces. Biophony, a key component of soundscapes, indirectly reflects regional biodiversity and ecosystem health. However, the complex relationship between vegetation structure and biophony in urban parks is not well understood. 

Methods: During the summer of 2024, we simultaneously collected high-resolution acoustic recordings and backpack LiDAR data from 52 sites in central Beijing parks. Six acoustic indices and 42 vegetation structural variables were calculated from these datasets. We used principal component analysis (PCA) and the XGBoost-SHAP model to identify and assess the importance of key vegetation variables influencing biophony. A generalized additive model (GAM) was then used to quantify the nonlinear relationships between these variables and biophony characteristics. 

Results: Our key findings are: (1) The power spectral density (PSD) of different biophony frequency bands showed distinct diurnal patterns. PSD at 2–4 kHz and 4–6 kHz exhibited similar circadian rhythms, while the 6–10 kHz band showed a staggered vocalization pattern. (2) Mean diameter at breast height (DBH), canopy relief ratio (CRR), and rumple index (RI) were key drivers of biophony across all frequency bands. Understory structure and canopy cover (CC) were dominant factors in regulating overall soundscape indices (ACI, ADI, and BIO). (3) Forest stands with a euphotic volume over 50% and medium-sized trees were more favorable for bird vocal activity. Increased CRR and canopy surface morphology significantly enhanced biophony, particularly insect sounds. (4) Excessive understory density and a large proportion of oligophotic volume were detrimental to the coexistence and propagation of multiple sound sources, which can reduce soundscape diversity. 

Conclusions: This study systematically reveals how three-dimensional vegetation structure influences biophony, identifies key structural factors for biophony patterns, and provides a scientific basis for soundscape optimization and biodiversity conservation in urban green spaces.

Key words: soundscape monitoring, acoustic index, LiDAR, urban forest, parks