SpatialSAVE

Spatial patterns in savannas may increase ecosystem resilience and reverse biome transitions

(funding from the European Union’s Horizon 2020 research and innovation programme under the 

Marie Skłodowska-Curie grant agreement No [101025056 ])

Savannas occupy about one-eighth of the land surface worldwide but the vegetation distribution in this biome is not properly understood. Furthermore, land use and climate change are propelling shifts in vegetation characteristics and their spatial distribution. Also, at the savanna-forest boundary, where the two biomes are known to exist as alternative states for same climatic conditions, the importance of spatial heterogeneity has been recognized but not well understood. This project will study how the dynamic interaction of vegetation with its environment leads to spatial organization of vegetation in the savannas including at the savanna-forest transition zone. Motivated by recent findings, it will further explore the significance of these patterned structures in increasing ecosystem resilience and reversing transition of ecosystem states.

Key objectives: 

(I) To understand spatial patterns at the savanna-forest boundary

(II) To determine the vegetation patterns in the savannas along the rainfall gradient.

(III) To investigate the resilience of the patterned states.

Open savannas can be classified into three broad ranges according to annual rainfall humid, mesic, and dry (blue bar indicates rainfall decreasing from humid to dry, from right to left). The humid range potentially demonstrates bistability with tropical forest. The mesic range shows possible bistability with savanna woodland, and the dry end shows possible bistability with barren desert (or grassland, which is not shown). The importance of fire (illustrated by the red shade in the fire bar) decreases from humid toward dry ranges, where water availability is the main limiting factor and driving force. The expected spatial structures are compared with symbolic Google Earth image examples. Source: Rietkerk et al. 2021, Science. 

Non-spatial models predict ecosystem tipping

Both in drylands and humid savannas, non-spatial models predict that ecosystems would collapse abruptly under increasing environmental stress. However, spatial patterns are important and when we consider spatial dynamics, the ecosystems seems to be much more resilient.

Humid savannas

Fig: Savanna-forest boundary (picture source: Huntley, B.J. (2023). The Guineo-Congolian Rain Forest Biome. In: Ecology of Angola. Springer, Cham. https://doi-org.proxy.library.uu.nl/10.1007/978-3-031-18923-4_12 )

Coexistence states at the savanna-forest boundary

The coexistence states where a part of the  landscape is in savanna state and the other half is in forest state can be observed at the savanna-forest boundary. Such states can exist beyond the classical tipping point thus rendering the ecosystem more resilient to climate change

 Drylands

Fig: Vegetation patterns in drylands (picture source: vegetation pattern formation. Ehud Meron. Physics Today, 2019 (72) 


                Front instability

Even before the tipping point is reached in the dry ecosystems, the boundary of vegetation patches can become unstable to form patterns such that there is no catastrophic shift.

               Turing before tipping

At the drier end where water limitation is the key driving force, there is onset of Turing pattern before the tipping point. On decreasing water availability, such patterns adapt and change wavelength and extends beyond the tipping point predicted by the non-spatial model.

Reference: