Main content area

Biome stability in South America over the last 30 kyr: Inferences from long‐term vegetation dynamics and habitat modelling

Costa, Gabriel C., Hampe, Arndt, Ledru, Marie‐Pierre, Martinez, Pablo A., Mazzochini, Guilherme G., Shepard, Donald B., Werneck, Fernanda P., Moritz, Craig, Carnaval, Ana Carolina
Global ecology and biogeography 2018 v.27 no.3 pp. 285-297
algorithms, climate, climate change, data collection, ecosystems, forests, grasses, habitats, models, pollen, prediction, shrubs, species diversity, trees, South America
AIM: The aim was to examine the links between past biome stability, vegetation dynamics and biodiversity patterns. LOCATION: South America. TIME PERIOD: Last 30,000 years. MAJOR TAXA STUDIED: Plants. METHODS: We classified South America into major biomes according to their dominant plant functional groups (grasses, trees and shrubs) and ran a random forest (RF) classification with data on current climate. We then fitted the algorithm to predict biome distributions for every 1,000 years back to 21,000 yr BP and estimated biome stability by counting how many times a change in climate was predicted to shift a grid cell from one biome to another. We compared our model‐based stability map with empirical estimates from selected pollen records covering the past 30 kyr in terms of vegetation shifts, changes in species composition and time‐lag of vegetation responses. RESULTS: We found a strong correlation between our habitat stability map and regional vegetation dynamics. Four scenarios emerged according to the way forest distribution shifted during a climate change. Each scenario related to specific regional features of biome stability and diversity, allowing us to formulate specific predictions on how taxonomic, genetic and functional components of biodiversity might be impacted by modern climate change. MAIN CONCLUSIONS: Our validated map of biome stability provides important baseline information for studying the impacts of past climate on biodiversity in South America. By focusing exclusively on climatic changes of manifested relevance (i.e., those resulting in significant habitat changes), it provides a novel perspective that complements previous datasets and allows scientists to explore new questions and hypotheses at the local, regional and continental scales.