Combined with a novel hybrid retrieval strategy, the method enables highly accurate estimation of canopy nitrogen content(CNC) and canopy phosphorus content (CPC) in diverse wetland ecosystems, from karst vegetation to mangrove forests.
Wetlands are among Earth’s most valuable ecosystems, vital for flood control, water purification, shoreline protection, and biodiversity conservation. Vegetation plays a central role, serving as both an ecological buffer and a sensitive indicator of environmental change. Nitrogen and phosphorus are essential nutrients driving plant growth, photosynthesis, and physiological adaptation. Accurate measurement of CNC and CPC provides key insights into wetland health. However, their subtle absorption features are often masked by pigments and water content in multispectral data, complicating efforts to monitor them. UAV-based hyperspectral sensors could resolve this issue, but their cost and complexity limit widespread application. The challenge has prompted researchers to explore spectral super-resolution and radiative transfer models as potential solutions for accurate nutrient retrieval.
A study (DOI: 10.1016/j.plaphe.2025.100059) published in Plant Phenomics on 3 June 2025 by Bolin Fu’s team, Guilin University of Technology, enables accurate and transferable estimation of CNC and CPC across wetland vegetation, offering a cost-effective solution for large-scale ecological monitoring.
The study introduced a constrained PROSAIL-PRO spectra-matching (CPSM) method that incorporates field-measured traits such as LAI, chlorophyll (Cab), and leaf water (Cw) to extend UAV multispectral reflectance into a continuous hyperspectral range (400–2500 nm). The method was validated through reflectance consistency checks against UAV data, derivative spectral comparisons with field spectra, and an improved Sobol sensitivity analysis to identify trait–wavelength influences. Results demonstrated that CPSM spectra closely matched UAV reflectance in both karst wetlands and mangroves (R² = 0.82–0.86; RMSE = 0.1091–0.0729), with relative errors below 5% and constrained errors below 10% for most pixels. The derived spectra also preserved key vegetation features such as the red-edge and water-absorption valleys (R² = 0.99 vs. measured; MRE = 4.58%), confirming reliability for canopy nitrogen (CNC) and phosphorus (CPC) estimation. Sensitivity analysis revealed LAI as the dominant driver (22.81–69.44% contribution at 400–684 nm), while pigments influenced reflectance only in the visible region, proteins responded weakly at 1433–2500 nm, and Cw dominated beyond 1300 nm. CPSM expanded sensitive subdomains for CNC and CPC to 1380–2400 nm and near the red-edge (701–725 nm), improving correlations by up to 0.41 and 0.64, respectively, with species-level peaks above 0.8 (CNC) and 0.7 (CPC). By integrating CPSM spectra with fractional-order derivative (FOD) processing and spectral indexes, researchers further enhanced correlations by 0.04–0.37. The new hybrid retrieval strategy (P4M + FOD ± Indexes) outperformed traditional models, reducing RMSE by 0.000200–0.000248 g cm⁻² and mean relative errors by 8.64–20.24%, achieving R² values of 0.46–0.98 and MRE as low as 5.91%. Robustness tests (LOOCV) confirmed stability even under small sample sizes, while adaptive elastic-net regression (AELR) consistently delivered the highest accuracy (CNC R² = 0.60–0.96; CPC R² = 0.72–0.98). Finally, the method successfully transferred to mangrove species (R² = 0.77–0.89; MRE = 9.65–16.87%), underscoring its strong potential for large-scale ecological monitoring.
The CPSM approach offers a cost-effective alternative to hyperspectral sensors for UAV-based ecological monitoring. By enabling accurate, transferable estimation of nitrogen and phosphorus across multiple wetland ecosystems, it provides a practical tool for biodiversity conservation, restoration projects, and carbon balance assessments. The method supports better-informed strategies for managing wetlands under climate change and contributes directly to global efforts toward achieving the UN Sustainable Development Goals (SDGs), particularly those related to life on land, clean water, and climate action.
###
References
DOI
10.1016/j.plaphe.2025.100059
Original URL
https://doi.org/10.1016/j.plaphe.2025.100059
Funding information
This research was supported by the National Natural Science Foundation of China (Grant number 42371341), the Natural Science Foundation of Guangxi Zhuang Autonomous Region (CN) (Grant number 2025GXNSFFA069008; 2024GXNSFAA010351), Key Laboratory of Tropical Marine Ecosystem and Bioresource Ministry of Natural Resources (Grant number 2023ZD02), Zhejiang Province "Pioneering Soldier" and "Leading Goose" R&D Project (Grant number 2023C01027).
About Plant Phenomics
Plant Phenomics is dedicated to publishing novel research that will advance all aspects of plant phenotyping from the cell to the plant population levels using innovative combinations of sensor systems and data analytics. Plant Phenomics aims also to connect phenomics to other science domains, such as genomics, genetics, physiology, molecular biology, bioinformatics, statistics, mathematics, and computer sciences. Plant Phenomics should thus contribute to advance plant sciences and agriculture/forestry/horticulture by addressing key scientific challenges in the area of plant phenomics.