Remote Sensing Monitoring And Analysis Of Driving Factors Of Spartina Alterniflora In Coastal Zone Of Guangxi

Coastal zone of Guangxi is vunerable to the invasion of Spartina alterniflora for sufficient space of tidal flats and suitable climatic conditions,causing reduction of living space of wetland vegetation,and posing serious threat to the stability of coastal wetland ecosystem.Clarifying the distribution and expansion trend of S.alterniflora is the basis for the management.Due to the lack of continuous and high-quality image data,remote sensing classification with single image can not effectively identify the actual coverage of S.alterniflora,thus limiting the understanding to the expansion of S.alterniflora.In the light of the problems above,the distribution of S.alterniflora in coastal zone of Guangxi from 2009 to 2020 were classified by random forest model,with phenological characteristics used as input features.The phenological characteristics were extracted from time series data of characteristic index,which were calculated based on Landsat 7/8 images,and fitted by harmonic analysis.Landscape pattern of S.alterniflora were analyzed by calculating the centroid of patches and landscape indices.The impacts of social and economic factors and environmental factors on the distribution of S.alterniflora were evaluated by structural equation model and Maxent species distribution model,respectively,with the analysis of the potential distribution of S.alterniflora accomplished at the same time.The resultes can provide scientific reference for controlling the expansion trend of S.alterniflora and improving management of wetland ecosystem in coastal zone of Guangxi.The main conclusions are as follows.(1)The extracted phenological features based on harmonic analysis can effectively express the growth information of S.alterniflora.The random forest algorithm combined with phenological features showed a higher accuracy in identifying the distribution of S.alterniflora,with the overall accuracy of the classification results in 2009 reached 97.33%,and the Kappa coefficient was 0.95.(2)From 2009 to 2020,the total area of S.alterniflora in coastal zone of Guangxi continued to increase,while the growth rate increased first and then decreased.The coverage of S.alterniflora in different regions showed different expansion stages.The distribution of S.alterniflora in Dandou Sea was affected by environmental conditions and human intervention,and reached saturation stage.The growth of S.alterniflora tended to saturation after continuous proliferation in tidal flat of Nanliu river estuary,while the growth of S.alterniflora in Tieshan Port and Yingpan Port was still in the rapid expansion period.(3)From 2009 to 2020,the centroid of S.alterniflora patches shifted from east to west,and patch shapes tended to be complex.Landscape distribution showed a trend of fragmentation,which was greatly disturbed by human activities The results showed that the analysis of structural equation model was credible,suggesting that aquaculture was the dominant social driving factor on the distribution of S.alterniflora in 2010 and 2020,while urbanization was the dominant social driving factor in 2015.(4)The critical environmental factors affecting the distribution of S.alterniflora were average sea surface temperature in the coldest season(18.67～20.01℃),average sea surface temperature in the warmest season(30.26～33.93℃),annual average sea surface temperature(>25.15℃),average precipitation in the wettest season(<178.59 mm),soiltype(preference in coastal saline soil)and slope.The total potential area of S.alterniflora was 14303.39 hm~2,which could be divided into high suitable habitats(9.87%),medium suitable habitats(71.56%)and low suitable habitats(18.57%).