关键词: 互花米草, 长时序遥感监测, 黄河三角洲, 湿地分类, 演化机理分析

Abstract: Aims In recent years, with the outbreak of invasive species—Spartina alterniflora, the structure and function of the wetland ecosystem in the Yellow River Delta Nature Reserve have been severely damaged. It is of great ecological value to analyze the invasion characteristics and landscape pattern changes of S. alterniflora in long time series. Methods In this study, we use long-term Landsat optical remote sensing images as the data source and takes the Yellow River Delta wetlands as the study area. Based on the support vector machine classification algorithm, we have obtained the wetland type information of the Yellow River Delta from 2000 to 2022. On this basis, we revealed the spatiotemporal dynamic changes of the wetlands. By combining 8 landscape pattern indices, the outbreak characteristics of S. alterniflora were studied and further analysis on the spatial pattern, evolution characteristics, and ecological factors of S. alterniflora were conducted. Important findings The results showed that: (1) The average overall accuracy of ground object extraction results obtained in 23 periods is 88.60%, and the average Kappa coefficient was 0.85. The Support Vector Machine algorithm exhibits higher classification accuracy than other traditional machine learning methods, and its results can well reveal the needs for analyzing the wetland evolution mechanism. (2) The results of wetland evolution over long time periods indicate that the expansion of S. alterniflora obvious phased characteristics: from 2002 to 2010, S. alterniflora expanded slowly; Since 2010, the expansion of S. alterniflora on both sides of the Yellow River estuary has entered an explosive period. By 2012, its area had increased to 22.63 km2; And it continued to grow steadily from 2012 to 2021. In 2021, the area of S. alterniflora had expanded to 50.42 km2. The landscape pattern index indicates that S. alterniflora shows high ecological advantage in the study area. The fragmented patches of S. alterniflora continuously gather and connect, gradually invading the native wetland vegetation and constructing a staggered spatial pattern with other vegetation communities. (3) The expansion of S. alterniflora in the study area shows a high correlation with hydrothermal conditions, sunshine duration, topography and mudflat environment, among which, there is a extremely significant correlation between the sunshine conditions, distance from the sea, and the growth of S. alterniflora. The research results of this article offer theoretical underpinnings for relevant departments in the rational utilization of wetlands in the Yellow River Delta Nature Reserve, as well as in the management and eradication of S. alterniflora.

Key words: Spartina alterniflora, long-term remote sensing monitoring, Yellow River Delta, wetland classification, evolution mechanism analysis