The Effects of Interannual Rainfall Variability on Tree–Grass Composition Along Kalahari Rainfall Gradient

Printer-friendly versionPDF version
Aug 15, 2017
Kailiang Yu, Michael Vijay Saha, Paolo D’Odorico



Precipitation variability has been predicted to increase in a global warmer climate, and is expected to greatly affect plant growth, interspecies interactions, plant community composition, and other ecosystem processes. Although previous studies have investigated the effect of intra-annual rainfall variability on plant growth and ecosystem dynamics, the impacts of interannual rainfall variability remain understudied. This paper uses satellite data and develops a new mechanistic model to investigate the response of tree–grass composition to increasing interannual rainfall variability in arid to sub-humid ecosystems along the Kalahari Transect in Southern Africa. Both satellite data and model results show that increasing interannual rainfall fluctuations favor deep-rooted trees over shallow-rooted grasses in drier environments (that is, mean annual rainfall, MAP < 900–1000 mm) but favor grasses over trees in wetter environments (that is, MAP > 900–1000 mm). Trees have a competitive advantage over grasses in dry environments because their generally deeper root systems allow them to have exclusive access to the increased deep soil water resources expected to occur in wet years as a result of the stronger interannual rainfall fluctuations. In relatively wet environments, grasses are favored because of their high growth rate that allows them to take advantage of the window of opportunity existing in years with above average precipitation and thus increase fire-induced tree mortality. Thus, under increasing interannual rainfall fluctuations both direct effects on soil water availability and indirect effects mediated by tree–grass interactions and fire dynamics are expected to play an important role in determining changes in plant community composition.

Read the full paper in Ecosystems.
Associated SESYNC Researcher(s): 
DOI for citing:
Share: Facebook Icon Twitter Icon Linked Icon