Speaker
Description
Forecasting ecosystem stability under climate change requires moving beyond precipitation totals toward a mechanistic understanding of how rainfall amount and timing, atmospheric demand, soil properties and biological legacies interact. This is especially critical in Mediterranean ecosystems, where annual plant communities experience strongly seasonal rainfall, high interannual variability and increasing warming-related atmospheric stress.
We synthesize 25 years of evidence from the Matta LTER station in Israel, one of the world’s longest-running rainfall manipulation experiments. The long-term record shows high resistance of aboveground herbaceous biomass to moderate drought: a 30% experimental rainfall reduction caused only a small biomass decline. This apparent stability indicates that cumulative winter rainfall alone is a poor predictor of ecosystem functioning. Instead, resistance is conditional on how water is delivered, stored and translated biologically into production.
To identify the mechanisms underlying this conditional resistance, we combined long-term monitoring with a new extreme multi-year drought experiment that reduced rainfall by 66% while altering the timing of the remaining rainfall. Water was delivered either at natural rainfall frequency, as many small frequent events, or as fewer large events separated by longer dry intervals. This design separated drought intensity from rainfall intermittency and revealed a soil-texture paradox: clay-rich Terra Rossa soils buffered drought magnitude by retaining water, but increased ecosystem sensitivity to the spacing of rainfall events. Prolonged dry intervals triggered functional reorganization, including strong declines in annual legumes and weakened demographic buffering through the soil seed bank.
Across the long-term record, machine-learning analyses further showed that vapor pressure deficit (VPD) and diurnal temperature range (DTR) can outperform rainfall totals as predictors of biomass production. Together, these results indicate a shift toward a compound-drought-limited state, where ecosystem stability is governed by interactions among hydrological supply, atmospheric demand, thermal variability, soil hydraulic buffering and biological legacies. This synthesis demonstrates the critical role of long-term Research Infrastructures (RIs) in detecting hidden ecological thresholds that short-term experiments or rainfall-based models may miss under climate change.
| Are you participating to the "AnaEE Environmental Rising Star Award "? | No |
|---|