Vegas, also called bofedales, are high Andean wetlands that develop under punishing conditions: extreme altitude, daily frosts and oxygen-starved soils. They originate where groundwater aquifers reach the surface in valley bottoms, depressions or slopes, feeding permanently irrigated plant communities. Their vegetation does more than simply grow there; cushion plants act as ecosystem engineers, retaining and regulating water flows and, combined with the lack of oxygen, promoting the accumulation of organic matter below the surface. This makes vegas the region’s main sequestered carbon stocks, biodiversity hotspots and water-regulating systems, providing fresh water for human populations and forage for native and domestic herbivores.

The research team, drawing on vegetation surveys of 50 vegas across the Argentine Central Andean Puna, focused on two floristic assemblages. The first is dominated by the cushion plant Oxychloe andina of the Juncaceae family, which forms dense and continuous vegetation tapestries typically found at stream headwaters. The second is dominated by Cyperaceae sedges, chiefly Eleocharis pseudoalbibracteata and Zameioscirpus atacamensis, which tend to occupy lower-altitude sectors often associated with increased salinity or salt flats. Together these groups covered 41 of the surveyed wetlands, distributed between 3,277 and 4,827 meters above sea level.

To model where suitable conditions for each assemblage exist, the researchers applied ecological niche modeling using n-dimensional minimum volume ellipsoids through the NicheToolBox package. Rather than modeling individual species, the approach targets groups of functionally associated plants, a strategy well suited to sparsely studied regions where many species have too few records to model alone. Working within the biotic-abiotic mobility framework, the team linked occurrence points at the center of each vega with bioclimatic variables from WorldClim 2.0 and altitude, selecting the least correlated variables to avoid redundant information.

The modeling procedure generated 120 candidate models for each group across three, five and seven environmental dimensions. For the Juncaceae vegas, the best model rested on three variables: mean annual temperature, diurnal temperature range and annual precipitation, with the niche centroid defined by values of 5.3 degrees Celsius, 18.5 degrees Celsius and 108 millimeters respectively. The Cyperaceae model required five variables, including altitude and temperature seasonality, with a centroid at 3,762 meters, a mean annual temperature of 7.2 degrees Celsius and pronounced seasonality of precipitation. The models achieved area under the curve values of 0.71 and 0.85 for the Juncaceae and Cyperaceae groups respectively, indicating reasonably reliable discrimination of suitable sites.

The results reveal a clear ecological split. Juncaceae-dominated vegas favor high, humid and cold environments, and their potential distribution concentrates toward the northeast of the Puna. The best model identified 1,439 vegas with habitat suitability above the applied threshold of 0.1, covering 15,418 hectares, or 28 percent of the total vega area in the study region, at an average altitude of 4,423 meters. Cyperaceae vegas, by contrast, occur at lower elevations with less humidity and higher temperatures, and their potential range extends further south. Their model encompassed 1,397 suitable vegas spanning 19,016 hectares, about 34.6 percent of the region’s vega surface, at a mean altitude of 3,922 meters.

Notably, 170 vegas showed suitability above the threshold in both models, reflecting the shared dominance of certain cushion species such as Oxychloe andina and Zameioscirpus atacamensis, and the gradual replacement of dominant species along the region’s environmental gradient. As humidity declines from the northeast to the southwest, communities shift from Oxychloe-dominated wetlands at the wetter end through the Cyperaceae assemblages in the middle toward halophyte-adapted communities at the arid extreme. Elevation and precipitation emerged as the master variables governing this transition, consistent with earlier field observations that Oxychloe andina thrives in high, wet, cold sites with shallow water tables, while Zameioscirpus atacamensis prefers saline conditions below roughly 4,200 meters.

The study also confronted the challenges of modeling in remote mountain deserts. Climate data in the high Andes are sparse, making bioclimatic layers derived from them imperfect, and the one-square-kilometer resolution of the environmental data may not capture the full altitudinal variation within small wetlands. The Juncaceae model showed a higher omission rate on test data, although the partial ROC test remained statistically significant. Interestingly, although vegas are often considered azonal ecosystems governed by local soil and groundwater conditions, the models demonstrate that their distribution is clearly tied to regional climatic gradients rather than strictly local factors.

The implications extend beyond biogeography. In a region where lithium mining, overgrazing, tourism and climate change are intensifying pressure on fragile water resources, knowing where each type of vega could exist provides a practical planning tool. Because different floristic assemblages differ in productivity, organic matter accumulation and carbon storage capacity, distinguishing their potential ranges helps conservation managers prioritize which wetlands to protect and where restoration might succeed. The authors frame this work as an initial step toward community-level niche modeling in the Argentine Puna, and they suggest that future models would benefit from incorporating groundwater depth, water chemistry and salinity, factors known to shape these communities alongside climate.

For a landscape where green ribbons of wetland mean the difference between desert and life, predicting where those ribbons can persist is more than an academic exercise. As Andean glaciers and precipitation patterns shift under a warming climate, maps of potential habitat for these engineer-dominated ecosystems may prove essential to safeguarding the water towers, carbon stores and biodiversity refuges of the high Puna.

Subject of Research: Ecological niche modeling of plant assemblages in high Andean Puna wetlands

Article Title: Modeling the potential distribution of floristic assemblages of high Andean wetlands dominated by Juncaceae and Cyperaceae in the Argentine Puna

Article References: Casagranda, E., & Izquierdo, A. E. (2023). Modeling the potential distribution of floristic assemblages of high Andean wetlands dominated by Juncaceae and Cyperaceae in the Argentine Puna. Vegetation Classification and Survey, 4, 47-58. https://doi.org/10.3897/vcs.95779

Image Credits: AI Generated

DOI: 10.3897/VCS.95779

Keywords: vegas, bofedales, high Andean wetlands, ecological niche modeling, Argentine Puna, Juncaceae, Cyperaceae, Oxychloe andina, carbon sequestration, biodiversity hotspots, climate change, lithium mining