Snapshot of a changing climate

By William Bond, Chief Research Scientist, SAEON

An overview of studies that have measured tree cover change in untransformed savanna areas. The green dots show sites (~ 10 km²) where tree cover has increased over time. The vast majority of savanna areas have experienced an increase in tree cover and bush encroachment (O’Connor et al., 2014; Stevens et al., 2014 – unpublished data)

View larger version

In our new, high CO₂ world it will be far harder to maintain open grassy systems than in the past (Picture: Mitzi du Plessis)

A rich African fauna, from elephants to migratory herds of springbok, has been replaced by cattle, goats and sheep (Picture: Johan Pauw)

Fire, an ancient ecological force, has been suppressed as roads, croplands and urban settlements have fragmented the landscape (Picture: Murray Ralfe)

In the late 1990s, South Africa made its first national analysis of potential climate change impacts on all key sectors of the country. As we are among the most biodiverse countries in the world, there were also analyses of climate change impacts on our flora and fauna.

The western half of the country was projected to get much hotter and drier. The predicted result for our biodiversity was alarming - near elimination of the succulent Karoo biome and of the northern half of the fynbos biome. These predictions were so catastrophic that potential climate change impacts on biodiversity began to be explored worldwide.

Now, nearly twenty years later, South African scientists have produced a report summarising actual changes in our biomes over the last few decades.

Photographic evidence of major ecosystem changes

Using photographic evidence, both ground-based and from aerial photography, and repeat surveys where available, researchers have documented major changes in our ecosystems. But the changes are not what were predicted.

The succulent Karoo has been remarkably stable and the fynbos shows little sign of dramatic changes, except for the major threat of invasion by alien trees. Instead, the really big changes are in our summer rainfall biomes, grasslands and savannas.

In the early years of the 20th century, South Africa was open veld with vast grassy landscapes. Today, trees and shrubs are thickening up in savannas, and dense thicket is invading grasslands. Many of our landscapes are no longer open grassy vistas, but have become covered with trees.

The process is widespread from dry country to our high rainfall mountain catchments. It is on such a large scale that you begin to wonder if there will be any grasslands left in a globally changing world.

Tall shrubs have also spread along streambanks and koppies in the Karoo. However, the major change is, ironically, the spread of grasslands into the eastern Karoo shrublands. The grass invasion is so complete in certain areas that grass-fueled fires are burning in places that never burnt before.

This invasion of shrublands by grasses is directly opposite to the predictions of John Acocks, an influential ecologist, in the 1950s. He argued that Karoo shrubs were invading grasslands, eating into our grazing resources because of overgrazing. The spread of grasslands westwards is also opposite to the predictions of the 1990s country study of climate change impacts on our biomes.

What was wrong with these analyses? What have we learnt and how can future projected changes be improved?

The problem, we think, lies in assuming that climate controls the distribution of life in our region with all the biomes neatly arranged according to rainfall and temperature. But not in South Africa, or, indeed, most of the African continent. Where you might expect forest, we have grasslands in our high Drakensberg mountains, while in the winter rainfall regions of the south-western Cape, fynbos dominates in the high rainfall ‘water towers’ of the mountains. Mingled in with the grasslands, or fynbos, are patches of forest, the alternative biome state.

In the early years of the 20th century, South Africa was open veld with vast grassy landscapes. Today, trees and shrubs are thickening up in savannas, and dense thicket is invading grasslands.

The grass is greener on the other side

In our dry country, you can also see profound differences in vegetation across a fence, with no change in climate at all. In our region, climate merely sets the potential vegetation. The actual vegetation depends, especially, on the history of fire and the activities of large mammals.

We have a wealth of long-term experiments that show the enormous importance of fire in our higher rainfall regions, and mammal herbivory in lower rainfall regions in shaping our vegetation. The implication is that land use, how people manage the land, is a major factor driving ecosystem change.

Fire, an ancient ecological force, has been suppressed as roads, croplands and urban settlement have fragmented the landscape. A rich African fauna, from elephants to migratory herds of springbok, has been replaced by cattle, goats and sheep. This means that the land user is a major factor in our changing terrestrial ecosystems.

Global factors are also at play

Global warming is now detectable in most of South Africa in temperature records from weather stations. There are not, as yet, widespread detectable trends in rainfall.

What is clearly also changing, however, is atmospheric CO₂, the invisible hand of global change. CO₂, of course, is the major greenhouse gas that is contributing to global warming. Since the early years of the 20th century, it has increased from about 300 parts per million (ppm) to 400 ppm due primarily to the use of fossil fuel. These are higher concentrations than have been experienced by plants for at least a million years. CO₂ has direct and indirect effects on plant growth. Plants use less water as CO₂ increases, so for the same rainfall, plants should grow more or have longer growing seasons than in the past.

Plants also benefit directly from CO₂, and capture more carbon by photosynthesis than in the past. Glasshouse studies on savanna trees have shown striking responses of some of our most common tree species which are contributing to large-scale woody thickening. Seedlings produce much larger root systems, packed with starch reserves, and even produce larger thorns and more chemical defenses as CO₂ increases. The ecological effect is that trees establish as seedlings more readily, survive fire and browsing as saplings, and grow into trees far more readily than in the past.


Trees and shrubs are thickening up in savannas and dense thicket is invading grasslands. The picture on the left was taken by Pole-Evans near Somerset East in June 1917. The picture on the right, of the same landscape, was taken by Hoffman and Masubelele in December 2009.

Model simulations and long-term experiments support the importance of CO₂ in contributing to the expansion of woody plants at the expense of grasses. Thus we have a new global change driver, particularly important in open ecosystems which have the climate potential to form forests but have been prevented from doing so by fire, herbivory and the people that manage the land.

In our new, high CO₂ world, it will be far harder to maintain open grassy systems than in the past. Management practices that your grandfather used effectively may be ineffective today.

Can we influence landcover change?

The good news is that, with ingenuity, we can influence future landcover change by developing new approaches to applying the familiar tools of mammal herbivory and fire. This option is not available in the cold countries of the north where undesirable ecological changes are the result of global warming which only major changes in greenhouse gas emissions can rectify.

The amount of carbon in the atmosphere matters! These are the roots of the common Acacia karroo (sweet thorn) exposed to different levels of CO₂. Relative to CO₂ levels typical of pre-industrial conditions (far left), saplings of common tree species exposed to the high CO₂ of the late 1990s (second from right) grew more than three times the biomass, making massive root systems with increased starch concentrations. These effects will promote rapid resprouting after fire and recovery from browsing. The increases in CO₂ are radically transforming growing conditions of trees (Bond and Midgley, 2012)

How will our changing landscapes influence the benefits we gain from nature?

The loss of grasses reduces livestock productivity, changes biodiversity of these systems, impinges on the tourist experience in game reserves (you cannot see animals in dense bush), and reduces water yield from streams and rivers. There are potential benefits too. For example, increased woody productivity is beneficial where trees are used for biomass burning, or where the wood is of value.

What of the future?

With our developing understanding of landcover change and its causes, we should be better able to develop management methods to direct trajectories of change to those most desirable for particular purposes.

As regards projecting future ecosystem changes, we know that climate-based projections will get it wrong. In our part of the world, you have to factor in land management and direct and indirect effects of CO₂.