Monitoring wetland plant diversity from space: Progress and perspective

• Published in: International journal of applied earth observation and geoinformation Vol. 130; p. 103943
• Main Authors: Sun, Weiwei, Chen, Daosheng, Li, Zhouyuan, Li, Saiqiang, Cheng, Siying, Niu, Xiaomeng, Cai, Yimeng, Shi, Zhenghua, Wu, Chaoyang, Yang, Gang, Yang, Xiaodong
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.06.2024; Elsevier
• Subjects: Diversity maintenance; Hyperspectral and multispectral images; Near infrared short wave; Spectral diversity; Spectral variation hypothesis; Taxonomic diversity; Diversity maintenance; Taxonomic diversity; Near infrared short wave; Hyperspectral and multispectral images; Spectral diversity; Spectral variation hypothesis
• ISSN: 1569-8432
• DOI: 10.1016/j.jag.2024.103943

•11,607 papers were published over the past 60 years.•Wetland plant diversity monitoring was not evenly distributed globally.•Near infrared shortwave length was the most reliable spectral region.•Hyperspectral data was better than multispectral data in inversing plant diversity.•Most studies focused on the taxonomic and within-habitat diversity. Wetlands are the one of ecosystems with the highest biodiversity, ecological service functions and carbon storage. Affected by the synergistic impacts of human activities and climate change, the global wetland area has decreased by 35 % since 1970, with far-reaching implications on biodiversity loss. Compared with manual ground investigations, remote sensing is considered to be the most promising method for monitoring wetland biodiversity change in order to formulate the effective conservation strategies due to its characteristics of non-contact detection, low cost and timely. Here we used bibliometric method to analyze the study sites, methods, conclusions and shortcomings of published papers globally over the past 60 years for wetland biodiversity monitoring. We show that global distribution of wetlands monitoring was uneven, mostly concentrated in the United States, China and Northern Europe. Current researches mainly focused on coastal, marsh and estuarine wetlands, while other wetland (e.g., lake wetlands, riparian wetlands, artificial peatlands and high-altitude and high-latitude peatlands) monitoring were still lacking. Overall, 20 remote sensing platforms and sensors were used, and the near infrared shortwave length (780 ∼ 1100 nm) was the most reliable and sensitive spectral region. Among various estimation methods, the accuracy of nonlinear, multi-independent variables, and hyperspectral remote sensing models were generally higher than those of linear, single-factor and multispectral models, respectively. The estimation accuracy was affected by both ground sampling time and plant phenology. Most studies focused on the taxonomic and within-habitat diversity (α-diversity) of single-layer communities (grassland), while few paid attentions to the functional and phylogenetic diversity of inter-habitat (β-diversity) and region (γ-diversity) in the multi-layer communities (forest and shrubland), and biodiversity-ecosystem functioning (BEF) relationships. We suggest that prospective studies should strengthen wetland plant diversity monitoring globally. The multi-dimensional spectral data are mined and fused to provide new monitoring models with high accuracy. The monitoring should focus on the scale effects (α, β and γ), BEF relationships, and the plant diversity change with environmental gradients.