The stability of savannas, past and future
- 21 Mar 2012
- 17:00 – 19:00 U.K. - England - London
- What time is this for me?
- Virtual
- Speaker details
- Prof. Steven Higgins Goethe University Frankfurt
- Event contact
- Eleonora Gandolfi (E.Gandolfi@southampton.ac.uk) Southampton
Abstract:
Large proportions of the Earth's land surface are covered by biomes dominated by C4 grasses. These C4 grasses originated during the Late Miocene, 3-8 Myr ago (Ma), but molecular and fossil evidence suggests that they evolved some 20 Ma earlier during the Early Miocene/Oligocene.
Explanations for this time lag invoke decreasing CO2 concentrations, and changes in the seasonality of climate and fire. However, there is still no consensus about which of these factors triggered C4 grasslands and savannas to invade forests.
We use a dynamic vegetation model, the aDGVM, to test how CO2, temperature, precipitation and fire as well as vegetation traits related to fire influence the potential for C4 expansion. Simulations are forced with Late Miocene climate simulations undertaken with the Hadley Centre coupled ocean-atmosphere-vegetation GCM, HadCM3L. We show that physiological differences between the C3 and the C4 photosynthetic pathway cannot explain C4 grass invasion into forests, but that fire is a crucial driver. Fire promoting traits in expand the environmental space where C4 dominated biomes can persist. We propose that C4 expansion could have occurred as a three-phase process where C4 grasses first replaced C3 grasses, then fire allowed C4 biomes to invade forests and finally, fire resistant savanna trees evolved and replaced fire sensitive forest trees.
We additionally use this model to examine how savannas may change to future climate and CO2 changes. We demonstrate that tropical grassland, savanna and forest ecosystems are likely to shift to alternative states.
Specifically, we show that increasing atmospheric CO2 concentration and climate change will force transitions to forest. The timing of these critical transitions is dependent on stochastic events and between site variance in the rate of temperature increase.