| 注册
首页|期刊导航|安徽农业科学|玉田水库消落带植物群落特征及多样性研究

玉田水库消落带植物群落特征及多样性研究

夏国勋 张海娜 鲁向晖 马露露 张利超

安徽农业科学2025,Vol.53Issue(5):49-55,63,8.
安徽农业科学2025,Vol.53Issue(5):49-55,63,8.DOI:10.3969/j.issn.0517-6611.2025.05.011

玉田水库消落带植物群落特征及多样性研究

Characteristics and Diversity of Plant Communities in the Yutian Reservoir Hydro-Fluctuation Belt

夏国勋 1张海娜 1鲁向晖 2马露露 1张利超1

作者信息

  • 1. 南昌工程学院水土保持学院,江西 南昌 330099
  • 2. 南昌工程学院水土保持学院,江西 南昌 330099||江西省鄱阳湖生态水利技术创新中心,江西 南昌 330099
  • 折叠

摘要

Abstract

The cyclic adjustment of water level and human activities in the hydro-fluctuation belt lead to frequent changes in plant community structure,and timely and accurate access to the process of plant community structure change in the hydro-fluctuation belt can provide technical support for the optimisation and adjustment of plant community structure.In this paper,the plant community of Yutian Reservoir in Jing-dezhen,Jiangxi Province was taken as the object of study,and the UAV remote sensing images were used as the basic data source,and ran-dom forest,decision tree,support vector machine and maximum likelihood methods were applied to classify the species of the plants in the hydro-fluctuation belt,so as to achieve the fine classification of the plant species in the hydro-fluctuation belt in the reservoir area by remote sensing,and to provide a scientific reference for the protection of plant diversity.The results showed that the classification accuracy of identif-ying the plant species in the hydro-fluctuation belt of the reservoir area using the random forest method could reach 86.48%,with the overall accuracy increased by 2.50 percentage points-5.74 percentage points and the Kappa coefficient increased by 0.03-0.07 compared with the other classification methods,and that the plant classification had strong integrity and high accuracy.The Simpson's dominance index of the plant communities in the hydro-fluctuation belt based on the field survey and random forest classification were 0.61 and 0.70,and the Shan-non-Wiener's diversity index was 1.01 and 1.26,respectively.Therefore,the use of unmanned aerial remote sensing imagery and random forest classification is an effective method for analysing the characteristics of the plant communities in the hydro-fluctuation belt of the reservoir area and for conserving the plant diversity.

关键词

消落带/无人机遥感/随机森林分类/植物多样性/植物识别

Key words

Hydro-fluctuation belts/UAV remote sensing/Random forest classification/Plant diversity/Plant identification

引用本文复制引用

夏国勋,张海娜,鲁向晖,马露露,张利超..玉田水库消落带植物群落特征及多样性研究[J].安徽农业科学,2025,53(5):49-55,63,8.

基金项目

国家自然科学基金项目(52269013) (52269013)

江西省水利厅科技项目(202224ZDKT17). (202224ZDKT17)

安徽农业科学

0517-6611

访问量0
|
下载量0
段落导航相关论文