Restoration should not prioritise trees or grasses but the ability of the ecosystem to switch between savanna and woodland
- Conserving Central India
- 5 days ago
- 2 min read
Grasslands and open forests (open natural ecosystems or ONEs) have been considered ‘wastelands’ since colonial times and targeted for restoration through tree planting. In this study, we use paleo-records to understand if Central India has a history of ONEs, and we find that the sites in our study region were often savanna in the last ten thousand years. This does upend the current understanding of the forests as a close canopy system that have recently been degraded into becoming an ONE.
Simultaneously, there has been a growing understanding that the grasslands in ONEs may be an alternative stable state, i.e. the system is a savanna in the presence of fire and herbivory and a woodland in the absence of these disturbance regimes. However, the bulk of the studies that suggest this are based on contemporary ecological data that are, at best, a hundred years old. Using paleo-data allowed us to examine the dynamics of these ecosystems at longer time-scales.

We find that, similar to contemporary studies, tree cover increases with increase in precipitation and decreases with an increase in fire. However, while we do often find patterns similar to alternative stable states (bimodality), a system where the tree cover decreases over time is able to return to its original tree cover if the disturbance regimes lasted for less than 300 years. This suggests that as long as seed-bearing trees are still alive, the system retains its ability to convert from grassland to one with high tree cover.
There is an ever-growing concern today for conserving grasslands as they may be very ancient and an alternative stable state. However, given that the system is able to switch between forest and savanna, it is unclear on what basis we should prioritise either forest or savanna. Instead, a better strategy may be to allow the system to switch between the states by ensuring that we are not unnecessarily preventing fire or herbivory, and by preserving niches with different species composition that may preserve the ability of the system to switch between states.
Citation: Agarwala, Meghna, Krishna Pavan Komanduri, Mohammad Firoze Quamar, Jayashree Ratnam, and Charuta Kulkarni. 2026. “Restoration Goals: Insights from Antiquity and Dynamics of Forest–Savanna Mosaics in Central India during the Holocene.” Ecological Applications 36(2): e70188.





This article about ecosystem restoration makes an important point about biodiversity, especially how healthy environments depend on balance rather than forcing one outcome over another. It reminded me of a difficult semester where I was balancing environmental studies and nursing coursework at the same time, and every deadline seemed to arrive together. I had hire someone to take my online Nursing exam to reduce one layer of pressure so I could focus more deeply on understanding the material. Sustainable systems really need realistic support structures to survive.