Distribution of African Forest and Savanna Biomes are Highly Predictable by Climate, Finds Study

Coupled field observations and phytoclimatic modeling show that the distribution of African forest and savanna ecosystems are highly predictable by climate, researchers report in a new study.

The findings suggest that the effects of climate change on the distribution of African forests and savanna may be more easily forecasted than previously recognized. An important yet challenging goal for ecological science is predicting how global vegetation patterns will be altered by ongoing climate change. Generally, the distribution of Earth's vegetation biomes is determined by climate.

However, at local or regional scales, multiple ecosystem types can occur under a single climate regime depending on past disturbances and species interactions. For example, across the global tropics, forest, and savanna biomes are both observed in regions that receive between 1000 and 2500 millimeters of annual precipitation.

This observation has led to the emergence of the alternative ecosystem state (AES) hypothesis, which states that qualitatively different AES ecosystems are self-maintaining and cannot be predicted from environmental forcing factors, such as climate.

To date, the AES hypothesis has been difficult to reconcile within ecosystem prediction models used to forecast the potential impacts of climate change. Here, Steven Higgins and colleagues evaluate how well a climate-based prediction model can predict the distribution of forest and savanna across tropical and subtropical Africa.

They compare predictions from a novel plant growth model and phytoclimatic transform, which combines plant species distribution data and climate system data to estimate the ability of climate to support different species in a given ecological niche, to a benchmark spatial dataset of defined savanna and forest sites across Africa. Higgins et al. show that the model is able to predict forest versus savanna occurrence with high accuracy.

According to the authors, the findings suggest that spatial variation in climate is the dominant influence on ecosystems at both large and relatively fine scales, challenging the view that climate alone cannot predict AES over large swaths of the Earth's surface.