A major impediment to understanding climate change-induced shifts in phenology arose from recent results that show observational and experimental studies make different predictions for future plant phenology. Specifically, in experiments, plant phenology is less sensitive to warming compared to what has been observed with recent warming trends. This discrepancy occurs across phenological datasets collected from around the world, but its underlying cause is unclear.
We test one possible explanation for the discrepancy: experimental warming causes soil drying, resulting in interactive effects of temperature and moisture on phenology. We examine if interactive effects of temperature and soil moisture in experiments lead to the apparent decreased sensitivity of plant responses to warming. This explanation arose because warming treatments dry soils, and drying can affect plant performance in ways that are divergent from effects of warming. We use newly compiled databases of daily experimental climate and corresponding plant phenology data to fit hierarchical models for phenological events, such as budburst day and leaf-out day.
We find that soil moisture is reduced by 0.2% per degree of experimental warming, with wide variation among experimental sites. As predicted, phenology was advanced in warmed plots, compared to control plots, and accumulated growing degree-days to bud-burst, leaf-out, and other phenological events were generally lower in warmed plots. After accounting for above-ground temperatures, phenological events occurred earlier in soils with higher water content. The negative relationship between soil moisture and phenology was present across diverse herbaceous plant species. This suggests that the drying caused by experimental warming may be responsible for the muted sensitivity to warming previously reported in experiments compared to observations. Additional future comparisons to observational data will allow a more complete understanding of the indirect, as well as direct, effects of applied treatments on phenology in climate change experiments. Forecasts of phenology under future climate change can therefore be improved.