Spatial Pattern And Changes Of Spartina Alterniflora With Different Invasion Ages In Yancheng Coastal Wetlands

As the expansion rate of S.alterniflora gradually slows down,more and more researches focus on studying its characteristic of different invasion ages and invasive stages,as well as its impact on the ecological environment.Analyzing the effect of S.alterniflora with different invasive ages on the ecological environment is important for comprehensively mastering the expansion rules of S.alterniflora,protecting biodiversity,and adapting to climate change.However,current studies on S.alterniflora with different invasive ages are all based on the comparative analysis of cross-sectional and sample survey data,mainly on the micro level,which are lack of quantitative analysis at landscape scale.Quantitative analysis at landscape scale can effectively reveal the spatial pattern of S.alterniflora with different invasive ages,thus further analyzing the change characteristics of S.alterniflora with different invasive ages and the effect of different invasion stages on the environment.Therefore,the analysis of the ecological processes and driving factors of S.alterniflora with different invasion ages at landscape scale that has important significance for understanding the invasion of S.alterniflora,and further revealing the relationship between coastal wetlands landscape pattern and ecological process.This stduy,combined with spatial pattern analysis of S.alterniflora with different invasive ages and its biomass at landscape scale,which is of great significance for revealing the effect of invasion ages of S.alterniflora on its biomass and the service function of local ecological system,as well as for formulating corresponding control strategies.Based on the Landsat OLI data,the supervised classification,decision tree classification and object-oriented method were used to map the landscape type in Yancheng coastal wetlands.Furthermore,the performance of various methods in mapping S.alterniflora was compared,and the characteristics of landscape pattern in the study area was analyzed.And then,based on the Landsat time series images and multi-class methods,the information of the actual distribution of S.alterniflora in the study area was extracted using the maximum entropy(MaxEnt)and the support vector machine(BSVM)methods.By testing the performance of the compressed time-series model,the effectiveness of the one-class classification methods(OCC)for S.alterniflora detection was evaluated.In addition,the distribution information of S.alterniflora for 20 years from 1996 to 2015 was extracted based on the OCC and visual interpretation methods,and the spatial distribution of S.alterniflora with different invasion ages in Yancheng coastal wetlands was mapped.Based on the principles of landscape ecology,the spatial patterns of S.alterniflora with different invasion ages in coastal Yancheng were analyzed.The main conclusions are as follows:(1)The object-oriented classification outperformed decision tree classification and supervised classification in terms of discriminative power.Meanwhile,all the methods exhibited better discriminability for mudflat and Phragmites australis,and yielded higher classification accuracy.However,the discriminability for roads and Suaeda salsa generated from those methods were relatively poorer.The object-oriented method not only can well reduce the "salt and pepper" effect of traditional pixel-based classification methods,but can greatly improve the classification accuracy.Thus,this method can provide important technical support and reference for the coastal wetlands classification based on the medium resolution remote sensing image.(2)Maxent and BSVM performed equally well,while Maxent appeared to have a more balanced performance over the summer months.April and December were deemed to be important periods for the detection of S.alterniflora.A compressed time-series analysis model,including only three variables(December NDVI,March Green and the third principal component in January,PC3),yielded higher accuracy than that of single-scene analyses,which indicated that time-series analysis can better detect S.alterniflora than single-scene analyses.The one-class classification method combined with a phenology-based detection strategy is therefore promising for the application of the long-term detection of S.alterniflora over extended areas.(3)The spatial pattern changes of S.alterniflora with different invasion ages have three stage characteristics:the patch area of S.alterniflora with invasion age of 15-20 years was very large with an average area of 425.40 hm2,formed in zones;the patch area of S.alterniflora with invasion age of 8-14 years became smaller with an average area of 153.64 hm2,and was formed in blocks with patchiness;the patch area of S.alterniforawith invasion age of 1-7 years was the smallest with an average area of 87.47 hm2.The seaward distance of S.alternifora increased by 917.78 m at a rate of 48.30 m/year,whereas the landward distance of S.alterniflora increased by 697.70 m at a rate of 36.72 m/year.Therefore,seaward expansion was the predominant S.alterniflora invasion pattern.The landscape shape index of the S.alterniflora with middle invasion ages was higher than those of lower and higher invasion ages.In conclusion,the patches of S.alterniflora distributed from the center to surrounding areas increased with invasion age.Higher S.alterniflora invasion ages showed more regular patch shapes and greater degrees of fragmentation.