Resources and disturbance predict sapling growth strategies in savannas

Background/Question/Methods

Our understanding of the dynamics of savanna remains imperfect, at least in part because ‘savanna’ can be used to describe a variety of systems.  Savannas can occur at low rainfall, where rainfall directly limits tree cover, or at higher rainfall, where closed forest canopies are possible but where disturbances from fire reduce tree cover.  Savanna tree demography and especially the establishment of savanna trees mediate this interaction of resource-limitation and disturbances from fire.  However, a general framework for understanding large-scale (i.e. continental) patterns in savanna tree establishment is so far lacking.

We have formulated a model to examine the effects of resource-fire interactions on optimal savanna tree establishment strategies. In reality, savanna trees employ a variety of strategies to deal with the combination of resource-limitation and disturbance from fire.  Persisting within the ‘fire trap’ by investing in storage below-ground is a common strategy; investing in height to escape the ‘fire trap’ is another.  However, rates of carbon assimilation are limited by limited resources, and plant strategies determine where plants with allocate those resources.

Results/Conclusions

We examine the effects of investment a) below-ground (i.e. in roots) and b) above-ground (i.e. in shoots or leaves) on lifetime reproductive success of individual savanna trees.  We find that the interaction between resource-limitation and disturbance produces surprising changes in expected sapling growth strategies along a resource gradient.  Comparison with available data on empirical patterns of sapling growth rates and sapling allocation strategies suggests that these results are plausible, although a more comprehensive evaluation of continental patterns of tree demographics is required.

These results suggest that understanding the types of dynamics that characterize savanna systems requires a consideration of more than just the drivers traditionally considered to determine tree cover in savannas – including rainfall, soils, fire, and herbivory. Plant functional type, or growth strategy, can also be expected fundamentally to determine the demographics of savanna trees and, by extension, the structure and tree cover of savanna systems.