Multispectral Sentinel‐2 Data Predicts Plant
CSR
Ecological Strategies: A Case Study From the Gandaman Wetland in Iran
Shahrbanoo Rahmani, Zeinab Jafarian, Ataollah Ebrahimi, Barat Mojaradi, Fabian Ewald Fassnacht ABSTRACT
Aims
This study aimed to extract the gradience of competitor, stress‐tolerators, and ruderals strategy types of plants in a wetland and examine the capacity of free multispectral Sentinel‐2 data to model and map these strategies.
Location
The study was conducted in the Gandoman Wetland in Iran.
Methods
We sampled 55 plots across the wetland and measured leaf area, fresh weight, and dry weight to compute the specific leaf area and leaf dry matter content of the plants to derive the competitor, stress‐tolerator, and ruderal strategies. Then, a partial least squares regression was used to model scores of these strategies that were derived from the plot‐level field data using Sentinel‐2A data.
Results
Our analysis accurately modeled the competitor, stress‐tolerator, and ruderal score with R 2 = 0.798, 0.665, and 0.545, respectively. A map of these strategies was derived by combining the predicted raster maps of competitor, stress‐tolerator, and ruderal values. The results corroborate previous findings based on hyperspectral data that these strategy types in wetland areas can be successfully mapped using multispectral Sentinel‐2 data.
Conclusion
The research discovered significant differences in ecological strategies among wetland plants in Iran, with competitive species in the center and ruderal species around the periphery. These findings agree with existing knowledge of the site and the known land use. Our study emphasizes the potential of multispectral satellite data to map competitor, stress‐tolerator, and ruderal strategy types to understand ecosystem functioning.