Abstract

The bromeliad Tillandsia landbeckii grows along the coastal scarp of the hyperarid Atacama Desert, one of the most extreme environments of the planet. A series of specialized adaptations such as unique trichomes, CAM metabolism and foliar morphology enable this plant to survive off scarce amounts of coastal fog that blow in from the ocean. By forming bands or “dunes”, these plants form complex ecosystems known as tillandsiales. Layered remains of these fog-specialist plants are often found buried and preserved within these dunes attesting to the establishment of these ecosystems over several millennia. In fact, the largest of these dunes have organic layers dated to >3200 cal yr BP (calibrated years before present). Here, we use stratigraphy, geochronology, stable isotopes (δ13C, δ15N), and biomass nutrient stoichiometry (C, N y P) from �ive different tillandsiales along a ~215 km fog gradient to establish how these dunes form and accumulate over time. We thus: a) establish a chronology of past dune establishment, b) propose an explanation for how these dunes accumulate, c) relate gradient changes in δ15N to present variations in fog intensity, and d) discuss variations in nutrient ratios over time. Across our gradient, modern tillandsiales evince a negative correlation between δ15N and fog intensity. We studied a dead Tillandsia dune system located along the southern margin of our gradient. Radiocarbon dates indicate that it “died” sometime shortly after 1200 cal yr BP. The low δ15N values compared to nearby living stands indicate that it lived under considerably greater fog input than today. These systems thus have the unique potential for studying regional variations in advective fog in northern Chile over the late Holocene.