Application of UAV images in monitoring flowering coverage and insect visiting activities in wetland plant communities
Application of UAV images in monitoring flowering coverage and insect visiting activities in wetland plant communities
复制标题
DOI：
10.11913/psj.2095-0837.2021.50467
复制
发表时间：
2021-10-01
期刊：
影响因子：
--
通讯作者：
Wang Xiao-Fan
中科院分区：
文献类型：
--
作者：
Deng Lu-Xi;Tang Sha-Sha;Wang Xiao-Fan
在群落水平的传粉生态学研究中。利用传统的调查方法对大比例尺的样地进行定量分析具有一定的技术局限性。这里.摘要探讨了利用无人机遥感影像数据在神农架大九湖亚高山湿地草本植物群落水平上研究传粉生物学的可行性。基于4个样地（1600 ~ 3000 m2）不同季节的无人机可见光遥感影像，采用支持向量机（SVM）分类方法计算不同颜色花卉的开花盖度。然后使用ContextCapture软件将其组合。结合16个样方（每个样方2 m × 2 m）内访花昆虫活动的野外调查。结果表明：（1）不同品种的开花盖度不同。在无人机图像中的彩色花朵与传粉者的数量显著相关。显示出指数关系。(2)随着无人机飞行高度的增加，开花盖度观测值呈下降趋势。(3)在不同的样地里。单位开花面积内传粉者数量无显著差异。在群落水平的传粉生态学研究中，使用传统的调查方法对大尺度样地进行定量分析存在一定的局限性，无人机遥感技术可能为此提供一种解决方案。
In studies on pollination ecology at the community level. the use of traditional survey methods for quantitative analysis of large-scale sample plots has certain technical limitations. Here. we explored the feasibility of applying unmanned aerial vehicle (UAV) image data to study pollination biology at the herbaceous plant community level in a subalpine wetland of Shennongjia Dajiuhu. The support vector machine (SVM) classification method was used to calculate flowering coverage of different-colored flowers based on UAV visible light images of four sites (1600 similar to 3000 m(2)) across different seasons. which were then combined using ContextCapture software. Combined with field surveys of flower.visiting insect activity within 16 quadrats (2 m x 2 m each). the results showed that: (1) Flowering coverage of different. colored flowers in the UAV images was significantly correlated with number of pollinators. showing an exponential relationship. (2) The observed value of flowering coverage showed a decreasing trend with the increase in UAV flight altitude. (3) In the different sample plots. the number of pollinators within a unit flowering area showed no significant differences. We also explored the feasibility of calculating flowering coverage through UAV images to monitor flowering season dynamics in the study area and estimate the number of pollinators.