无人机的出现,给生态调查带来关键性革新。而使用无人机进行生态调查,植物遥感分类是关键,基于平地的无人机植物物种分类创新运用于边坡,使用可见光正射影像联合nDSM(normalized digital surface model,归一化数字表面模型)对边坡植物...无人机的出现,给生态调查带来关键性革新。而使用无人机进行生态调查,植物遥感分类是关键,基于平地的无人机植物物种分类创新运用于边坡,使用可见光正射影像联合nDSM(normalized digital surface model,归一化数字表面模型)对边坡植物物种进行分类。结果表明,边坡样地的分类的精度达85%,自然样地达84%,与没有加入nDSM的分类结果对比,边坡、自然样地分类精度分别增加了32%和16%。在边坡条件下可见光正射影像与nDSM结合,可大幅度提升边坡植物物种分类精细度。展开更多
Environmental heterogeneity significantly affects the structure of ecological communities.Exploring vegetation distribution and its relationship with environmental factors is essential to understanding the abiotic mec...Environmental heterogeneity significantly affects the structure of ecological communities.Exploring vegetation distribution and its relationship with environmental factors is essential to understanding the abiotic mechanism(s)driving vegetation succession,especially in the ecologically fragile areas.In this study,based on the quantitative analysis of plant community and environmental factors in 68 plots at 10 different transects in the Minqin oasis-desert ecotone(ODE)of northwestern China,we investigated desert vegetation distribution and species-environment relationships using multivariate analysis.Two-way indicator species analysis(TWINSPAN),detrended correspondence analysis(DCA),and canonical correspondence analysis(CCA)methods were used.A total of 28 species,belonging to 27 genera in 8 families,were identified.Chenopodiaceae,Zygophyllaceae,Gramineae,and Leguminosae were the largest families.Annual and perennial herbs accounted for 28.60%of the total number of plants,while shrubs(42.90%)were the most dominant.Nitraria tangutorum was the constructive species of the desert plant community.We divided the 68 plots surveyed in this study into 7 community types,according to the results of TWINSPAN.The distribution of these 7 communities in the DCA ordination graph showed that species with a similar ecotype were clustered together.Results of CCA indicated that groundwater was the dominant factor influencing vegetation distribution,while distance between plot and oasis(Dis)and soil electrical conductivity(EC)were the local second-order factors.Our study suggests that optimizing the utilization of groundwater in oases is key to controlling the degradation of desert vegetation.The favorable topographic conditions of sand dunes should be fully utilized for vegetal dune stabilization,and the influence of soil salinity on the selection of afforestation tree species should be considered.展开更多
文摘无人机的出现,给生态调查带来关键性革新。而使用无人机进行生态调查,植物遥感分类是关键,基于平地的无人机植物物种分类创新运用于边坡,使用可见光正射影像联合nDSM(normalized digital surface model,归一化数字表面模型)对边坡植物物种进行分类。结果表明,边坡样地的分类的精度达85%,自然样地达84%,与没有加入nDSM的分类结果对比,边坡、自然样地分类精度分别增加了32%和16%。在边坡条件下可见光正射影像与nDSM结合,可大幅度提升边坡植物物种分类精细度。
基金supported by the National Key Research and Development Program of China(SQ2016YFHZ20617-03,2018YFC0507102-05)the National Natural Science Foundation of China(41661008,41761051,41761006,41661064,31560128,41801102)
文摘Environmental heterogeneity significantly affects the structure of ecological communities.Exploring vegetation distribution and its relationship with environmental factors is essential to understanding the abiotic mechanism(s)driving vegetation succession,especially in the ecologically fragile areas.In this study,based on the quantitative analysis of plant community and environmental factors in 68 plots at 10 different transects in the Minqin oasis-desert ecotone(ODE)of northwestern China,we investigated desert vegetation distribution and species-environment relationships using multivariate analysis.Two-way indicator species analysis(TWINSPAN),detrended correspondence analysis(DCA),and canonical correspondence analysis(CCA)methods were used.A total of 28 species,belonging to 27 genera in 8 families,were identified.Chenopodiaceae,Zygophyllaceae,Gramineae,and Leguminosae were the largest families.Annual and perennial herbs accounted for 28.60%of the total number of plants,while shrubs(42.90%)were the most dominant.Nitraria tangutorum was the constructive species of the desert plant community.We divided the 68 plots surveyed in this study into 7 community types,according to the results of TWINSPAN.The distribution of these 7 communities in the DCA ordination graph showed that species with a similar ecotype were clustered together.Results of CCA indicated that groundwater was the dominant factor influencing vegetation distribution,while distance between plot and oasis(Dis)and soil electrical conductivity(EC)were the local second-order factors.Our study suggests that optimizing the utilization of groundwater in oases is key to controlling the degradation of desert vegetation.The favorable topographic conditions of sand dunes should be fully utilized for vegetal dune stabilization,and the influence of soil salinity on the selection of afforestation tree species should be considered.
