Application Of UAV Images In Monitoring Flowering Coverage And Insect Visiting Activities Of Wetland Plant Communities

In studies of pollination ecology at community level,with the expansion of the research area and monitoring scale,long-term and multi-spatial-scale research work is usually required.There were certain technical limitations in using traditional survey methods for quantitative analysis of large-scale sample plots.At the same time,the research and construction of pollinator networks need to acknowledge and monitor the interaction between plants and pollinators,and to study the interaction between plants and pollinators and the relationship between plants from the community level is conducive to clarifying the relationship between pollinators and plants.As well as discussing the role of pollinators in the community,more and more attention has been paid to the changes in the number of pollinators and the potential impact on the pollination services of crops and wild plants.With the development of remote sensing technology,near-surface remote sensing technology has brought new opportunities for pollination ecology research.To explore the feasibility of applying unmanned aerial vehicle(UAV)image data to pollination biology at community level,this study adopted herbaceous plant community of the subalpine wetland Shennongjia Dajiuhu as objects,and the classification method of Support Vector Machine(SVM)to calculate flowering coverage of different-colored flowers,by utilizing consumer UAV to obtain the visible light image data of 4 sites in sizes ranged from 1600 m~2 to 3000 m~2 each in different seasons,which are then joined together by software.When combined with the field survey of flower-visiting insect activities within 16 quadrats of 2m x 2m each,the results of data analysis showed:(1)In the natural environment,the number of flowering(R2=0.527,P<0.01)and the visual coverage of flowering(R2=0.855,P<0.01)both have a significant positive correlation with the flowering coverage based on UAV images;(2)With the increase of UAV flight altitude,the observed value of flowering coverage was trending downward;(3)The proportion of individuals in the functional groups of flower visiting insects varied with the monthly variation of flowering coverage;(4)The number of pollinators within a unit flowering area in different colors had no significant difference,and the flowering coverage of different-colored flowers shown in UAV images had significant correlation with the number of pollinators,which presents as an exponential relationship(R2=0.678,P<0.01).This study also discussed the calculation of flowering coverage through UAV images so as to monitor the seasonal dynamics of flowering and estimate the number of pollinators in the study region.