Floristic types of high-Andean wetlands from northwest Argentina and their remote-sensed characterization at a regional scale

Aims High-Andean vegas are key functional wetlands in the Puna ecoregion. Plant communities in combination with ecogeographic characteristics determine their functional processes. In this study, we identified groups of vegas based on their plant composition and characterized these groups with spatial and spectral variables representing their ecogeographic context. Location Argentine Puna and High-Andean ecoregions. Methods We recorded the species composition and cover of plants in 50 vegas distributed along a ecogeographic gradient. We calculated six spatial and 14 spectral variables for each vega. We performed a correspondence analysis (CA) to explore species data and used the site's scores in a k-means analysis to identify groups of vegas. Then, we characterized each group of vegas with spatial and spectral variables with the v.test using the 'catdes' function in the FactoMineR package. Results The CA showed five groups of vegas segregated by the plant species composition. Each group was related to different spatial and spectral variables showing an ecogeographic gradient. Vegas with Poaceas were located at higher altitude and lower latitude and longitude (Group 1, Festuca nardifolia and Deschampsia hackelii). Vegas dominated by cushion species had higher humidity (Group 2, Oxychloe andina), and higher and more stable productivity (Group 3, Eleocharis pseudoalbibracteata), while vegas with halophytic species were associated with a larger area, higher salinity, and lower humidity (Group 4, Amphiscirpus nevadensis), and lower productivity (Group 5, Lycium humile and Salicornia pulvinata). Conclusions Our results are the first floristic classification and remote-sensing characterization of high-Andean vegas at a regional scale. This information shows the variation of these ecosystems and suggests that remote sensing, complemented with field information, could help to identify types of vegas at regional scales. This information is relevant for land planning and sustainable management of these key ecosystems in the context of threats of global change.