Small, gnarled trees, some with bark charred by the passage of fire, stand in the midst of a grassy carpet. Anyone who has seen it before will immediately recognize the Cerrado, the Brazilian savannah. In Africa and Australia, the other two continents that feature this biome, the savannahs present very similar landscapes. But the similarity is only superficial, because the Cerrado has greater biodiversity, to the extent that it is listed among the 34 areas having the world’s greatest abundance of species threatened with extinction, or hotspots.
New information indicates that the savannahs on the three continents also differ in the way they respond to fire, moisture and temperature, according to an international group that includes Brazilian participants. The group’s findings were reported in the January 2014 issue of the journal Science, based on data compiled during more than 100 studies conducted at 2,154 savannah sites in South America, Africa and Australia. In addition to their importance for our understanding of how savannah environments operate, the findings are essential for the development of models that will predict savannahs’ reaction to climate change and estimate their ability to mitigate that change by removing carbon from the air.
“We have seen the role that evolutionary history apparently played in shaping the contemporary dynamics of the biome,” says Caroline Lehmann of the University of Edinburgh, in Scotland. This broader perspective is, in her view, the most striking finding of the work she coordinated. The differences apparently arose because the savannah is relatively young, having emerged between three and eight million years ago. By then the continents had been separated for some time, and their flora and fauna had accumulated striking differences. The tree species, predominantly Myrtaceae (a family that includes pitanga–or Surinam cherry–, guava, jabuticaba, and eucalyptus) in Australia and Leguminosae in Africa, are distinct in their phenology—the periodicity with which they produce flowers and fruits—and in fire resistance, growth rates, and architecture. The Cerrado, the most diverse of the savannahs, does not have a predominant botanical family.
A closer look at the environmental factors that govern these ecosystems revealed that they underlie functional differences. In Africa and Australia, rainfall and temperature have a strong effect on increasing fire frequency, because they are conducive to the growth of grasses. To a lesser extent, these factors also influence tree size. In South America these relationships are very weak, both in Brazil and in Venezuela, where savannah vegetation can also be found. The variation from one continent to another surprised the researchers, who expected to see greater homogeneity. “In retrospect, it seems quite obvious when you consider the diversity in the architecture and phenology of the trees in these regions,” Lehmann says.
The important point is that this variation means it is not possible to use a single model to predict tree biomass under certain environmental conditions, for example, or how the vegetation will react to changes in global temperature. One special feature of the Cerrado is that it evolved in a moister environment than the other savannahs. “The same climate that features Cerrado here would have forests on the other continents,” says forest engineer and study co-author Giselda Durigan of the Forestry Institute of São Paulo State in Assis, a municipality in the state’s interior.
The particular features of Africa are also a result of the wide variety of large herbivores—such as elephants, antelopes and zebras in heavily populated herds—whose voracious appetite for plants prevents the survival of young tree shoots and creates a landscape where predominantly grassy fields are much more common. “The absence of megafauna in South America is largely responsible for the diversity of the Cerrado,” Durigan notes.
Without large herbivores—here often represented by cattle—what gives the Cerrado its open physiognomy is fire. When there are no fires, trees grow, proliferate and inhibit germination and development of endemic species that are not shade-tolerant. Without fire or grazing, the grass itself can be harmful to seedlings that require light. An example of how fauna and fires are an integral part of the ecosystem appeared in the research that Durigan is conducting at the Santa Barbara Ecological Station in inland São Paulo State. She found a plant less than 10 centimeters high and discovered it was a specimen of Galium humile, from the coffee family, a species that had not been collected in the state since 1918. Curiously, the discovery occurred precisely in an area that in recent decades was heavily subject to fires and grazing. “The flora and fauna of the Cerrado depend on fire,” Durigan points out. “In Brazil we are going to have to learn to use it as a management tool, now that the law provides for that practice for the good of the ecosystem.”
Research studies such as the one by Durigan’s group were the basis for the article published in Science, which brings together data from many other research groups. “It is the kind of study that expands the scope but sacrifices details,” Durigan comments. She was invited to the meeting in Australia that formed the working group in 2009, but was unable to participate due to a scheduling conflict that sent her to a different event in that same country. As a result, the only Brazilian representative was forest engineer Jeanine Felfili of the University of Brasília (UnB). But shortly afterwards, Felfili suffered a stroke and died, and her contribution was carried out in part by Ricardo Haidar, who at the time was her master’s degree student. Nevertheless, the journal rejected an initial version of the paper in 2013 because it lacked sufficient data from South America. Lehmann then called on Durigan, who at the time was not only available but had also just finished participating in an extensive survey of the Cerrado and had all the necessary data in her head and on her computer. “Many of the data were in articles written in Portuguese or even in theses,” she notes. In practice, therefore, they were invisible to foreigners.
Eyes on the future
With Durigan’s contribution, the study became more representative, employing more robust statistical models to estimate the effect of each variable on the biomass of the savannah. These models are also used to predict what might happen to the size of the savannahs in the event of the climate changes predicted for the coming decades. Under the scenario of an increase of four degrees Celsius (°C) in the average annual temperature, the study showed striking differences between the global and regional models of biomass change in the savannahs. In Africa, for example, the model that does not distinguish between continents predicts a slight reduction in biomass, while the specific model indicates there will be an increase. Under the same scenario for South America, the regional model predicts a much greater reduction in biomass than what is predicted by the global simulation.
“The biomass prediction maps derived from our statistical models are adequate for illustrative purposes,” Lehmann says in perspective. “But in fact, people exert an enormous influence on present-day biomass patterns through deforestation, agriculture, fishing and selective clearing.” She therefore conjectures that there is a notable mismatch between the predictions generated by the models and what is really happening. And she points to the Cerrado, which has undergone much more extensive transformations than the other savannahs due to its use for agriculture, and has already lost nearly half its territorial expanse.
But it is still impossible to anticipate what the environmental changes will bring about in the savannahs, and not just because of uncertainty about what will happen in each continent’s climate. The problem is especially complex for these ecosystems because of their enormous diversity within and across continents. The study focused on the most typical savannahs, which are more or less equally divided between trees and grasses. But on each continent the biome can range from dense grasses to a denser forest of tall trees with a sparse herbaceous layer. “The increase in concentrations of atmospheric CO2 will likely affect tropical grasses and trees differently and change the competitive balance between these key plants of the system,” Lehmann explains. The effects will vary according to region. “I can say that our lack of understanding of how the savannah systems might respond to climate change is a critical knowledge gap that should be taken seriously.” She believes that savannahs, which cover about 20% of the Earth’s surface, should be studied as diligently as are the Amazon and other tropical forests.
Intrigued by the weak relationship between temperature and rainfall variations and vegetation in the Cerrado, Durigan believes she will find answers beneath the surface. The physical characteristics of the soil have a major influence on the availability of water for plants, which need these reserves to withstand periods of drought. “When the soil is clayey, plants respond to a four-month drought as if it lasted only two months,” she explains. This occurs because clay can retain water longer and in higher quantities than can sand. “But when there is too much clay, water is retained in such a way that plants are unable to capture it.” The ideal conditions for plant development, therefore, involve a subtle balance among soil components that is more variable from one point to another in the Cerrado than in the other savannahs.
The models produced in the Science study to examine the relationship between environmental factors and tree biomass took into account the percentages of carbon and sand in a soil layer 50 centimeters deep. Carbon serves as a measure of the organic matter or nutrient content in the soil, and sand content provides an estimate of the soil’s water retention capacity. But these indicators—chosen because they are available for savannahs throughout the world—are insufficient, according to Durigan.
In giving indications of environmental variables that are important for the savannahs, the study suggests important directions for future research efforts. Durigan visualizes what would be needed to gain a better understanding of the complex relationship between soil, climate and the Cerrado: a research network with groups working throughout the biome, digging trenches of various depths in order to examine the soil and relate its properties to the size and other features of the vegetation.
LEHMANN, C. E. R. et al. Savanna vegetation-fire-climate relationships differ among continents. Science. v. 343, n. 6.170, p. 548-52. 31 jan. 2014.