You can't have your cake and eat it too: thoughts on savanna palaeobiology

By Alastair Potts

Dr Fred Kruger (University of Free State) published a fascinating article entitled "Palaeobiology of the South African savanna and lessons for modern ecologists". For anyone with an interest in savanna ecology, I suggest you read it.

This article touches on a combination of my interests: it was the natural history of the lowveld savanna that fascinated me as a child (and ultimately led to a career in botany), and I am starting to pay more attention to how we reconstruct the past environmental changes.

The focus of the article is a discussion about what insights palynology can provide for understanding modern African savannas. Ultimately, the narrative is one of regional stability where the modern-day tree flora has persisted for the last 300,000 years, but that there have been many reorganisations of the savanna communities. Although local extinctions have occurred, regional extinction of a species is very rare. Thus, savanna communities can be viewed as being loosely assembled with "compositional oscillation".

So, where is the cake? Dr Kruger aptly argues that the palynological record provides ecologists with the opportunity to "become comfortable with thinking in evolutionary terms", and that contemporary savannas can only be properly understood with knowledge of their pasts. But there is another theme that is brought up: that of reconstructing past environments using the same palynological data using a "biomization" procedure. This procedure creates links between current environments and communities. But can community information, based on loosely assembled species, also be used to reconstruct past environmental changes?

If we look at the figure below from Staver et al. (2011),  % tree cover (an indicator of savanna or forest; forest in this case defined as >55%) of Africa quickly loses any relationship with mean annual precipitation. It is only in the driest states where there is any predictability.

Given such patterns, Staver et al. (2011) suggest that forest and savanna are alternative stable states that can occur within the same climate. Different types of savannas are also likely to occur within the same environment. Admittedly, there may be specific species that will be able to improve the resolution of the species-environment link. However, species (and thus communities) in savannas are affected by far more than just climate – as discussed in the article – such as fire history and herbivory.

A range of palynological studies reconstructing savanna environments have stated that within the profiles is evidence of climatic changes in temperature and precipitation (e.g. cooling and drying trends). I will immediately admit that I have not read most of these palynological papers – as they have been outside my geographical area of interest – but they are now on my reading list. However, what struck me from this article is that the same data are being used to make two, possibly mutually exclusive points: 1) savanna community assembly is transient and not strongly related to climate, 2) palaeoclimate reconstructions are possible assuming that communities are strongly related to climate.

The title of Dr Kruger's article suggests that ecologists can learn a lot from palaeobiologists. I suspect that palaeobiologists may have more to learn from ecologists.


References

Kruger, F., 2015. Palaeobiology of the South African savanna and lessons for modern ecologists. Transactions of the Royal Society of South Africa 70, 117-125.

Staver, A.C., Archibald, S., Levin, S.A., 2011. The global extent and determinants of savanna and forest as alternative biome states. Science 334, 230-232.