Early Invasion Process Monitoring of Spartina Alterniflora Using Long Time Series High-Resolution Satellite Remote Sensing Imagery

<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Spartina alterniflora (S. alterniflora)</i> rapidly invade new habitats by seed jumping propagation, and the early invasion process is characterized by discrete large patches and diffuse distribution small-scale patches. Monitoring these early-invasion-stage patches is very important for effectively stopping the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">S. alterniflora</i> invasion trends. In this study, a long time series of high spatial resolution (0.5 and 1 m) satellite remote sensing images from 2010 to 2020 were used to monitor the early invasion period of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">S. alterniflora</i>. Using remote sensing technology, the early invasion behavior of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">S. alterniflora</i> at patch scale was monitored and analyzed in the Yellow River Delta with the object-oriented image segmentation and support vector machine classification methods. The results indicate that the discrete patches were incorporated into bigger patches or gathering regions within a few years: 56.13% of the independent large patches lasted only for 1 year, and only 0.25% of which can last for 10 years. Second, we newly found out that the newborn <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">S. alterniflora</i> seedlings are uniformly distributed between discrete large patches and then quickly aggregate into a large gathering area within the next 2 years. Third, for the area that was undergoing invasion, the mean <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">S. alterniflora</i> coverage rate was stable at approximately 50%. Fourth, dominant factors that determined the invasion trend of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">S. alterniflora</i> are the elevation of the intertidal zone, the tidal channel distribution and the topographic changes caused by <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">S. alterniflora</i> invasion.