The invasive plant Mikania micrantha Kunth ( M. micrantha ) from South America poses a significant threat to the stability and biodiversity of ecosystems. However, an effective and economical method to control M. micrantha is still lacking. RNA interference (RNAi) has been widely studied and applied in agriculture for trait improvement. Spray-induced gene silencing (SIGS) can produce RNAi silencing effects without introducing heritable modifications to the plant genome and is becoming a novel nontransformation strategy for plant protection. In this study, the genes encoding chlorophyll a/b-binding proteins were selected as targets of RNAi, based on high-throughput sequencing of M. micrantha transcriptome and bioinformatic analyses of sequence specificity. Three types of RNAi molecules, double-stranded RNA, RNAi nanomicrosphere, and short hairpin RNA (shRNA), with their corresponding short interfering RNA sequences were designed and synthesized for SIGS vector construction, from which each RNAi molecule was transcribed and extracted to be sprayed on M. micrantha leaves. Whereas water-treated control leaves remained green, leaves treated with RNAi molecules turned yellow and eventually wilted. Quantitative real-time PCR showed that the expression levels of target genes were significantly reduced in the RNAi-treated groups compared with those of the control, suggesting that all three types of RNAi herbicides effectively silenced the endogenous target genes, which are essential for the growth of M. micrantha . We also found that shRNA showed better silencing efficiency than the other two molecules. Taken together, our study successfully designed three types of RNAi-based herbicides that specifically silenced endogenous target genes and controlled the growth of M. micrantha. Moreover, we identified a gene family encoding chlorophyll a/b-binding proteins that is important for the growth and development of M. micrantha and could serve as potential targets for controlling the spread of M. micrantha.

Abstract. State-of-the-art point cloud classification methods mostly process raw point clouds, using a single point as the basic unit and calculating point cloud features by searching local neighbors via the k-neighborhood method. Such methods tend to be computationally inefficient and have difficulty obtaining accurate feature descriptions due to inappropriate neighborhood selection. In this paper, we propose a robust and effective point cloud classification approach that integrates point cloud supervoxels and their locally convex connected patches into a random forest classifier. We apply a centroid cloud extracted from supervoxels into the proposed classifier, which effectively improves the point cloud feature calculation accuracy and reduces the computational cost. Considering the different types of point cloud feature descriptions, we divide features into three categories (point-based, eigen-based, and grid-based) and accordingly design three distinct feature calculation strategies to improve feature reliability. The proposed method achieves state-of-the-art performance, with average F1-scores of 89.16%, respectively. The successful classification of point clouds with great variation in elevation also demonstrates the reliability of the proposed method in challenging scenes to some extents.