Much needed simplification

As a result, the entire methodology for establishing carbon levels has become simplified. Indeed the model developed by Jean-François Bastin and his colleagues has now made it possible to establish carbon levels with a margin of error of less than15% by measuring only 5% of the trees present! “If the objective is to regularly measure and verify the state of forests over large areas, we can now recommend only measuring trees that are 50 centimetres (to 1.30 metres in height), which represents barely 7% of the trees currently measured. Those trees with a diameter in excess of 50 centimetres explain more than 90% of the variation in carbon stocks. The mass of these trees is so important to the balance that the rest of the measurements become virtually useless in the context of carbon levels (it is clear that it would be different if the dynamics or floristic composition of the forest were studied)”.

The second major achievement of the study consists in providing an answer to the question: “What are these large trees”?  And in particular: “are they largely distributed across central Africa”? “A year ago, in the Amazon basin, research concluded that the main structuring elements of the forests were the large trees, showing that around 1% of the species present contain more than 50% of the carbon stocks. These species, which make a disproportionate contribution to the carbon levels are known as “hyperdominant”. Our work shows that the situation is slightly different in Africa. While a very small proportion of species listed (1.5%) also contain the major part of the forest carbon stocks, we noticed that a non-negligible proportion of these “hyperdominant” species are not particularly large: only rarely do they exceed 70 centimetres in diameter (to 1.30 metres in height). These so-called “hyperdominant” species are hyperdominant not with regard to their size but to their abundance and their wide distribution in the Congo basin. These species therefore constitute the key to understanding the carbon levels”.

These results provide attractive and relatively short-term possibilities for forest conservation programmes. Still today, in order to effect mapping of carbon levels and identify the risk zones and zones of interest in terms of conservation, methods which are said to be “indirect” are used. These methods are based on information from satellites (light reflection by vegetation, average height of the canopy), which do not directly constitute a measurement of carbon levels. In the future we will undoubtedly be able to do a lot better: “Because we now know that most of the information is contained in the largest trees and that these are directly observable in satellite images, it will probably be possible to extract “direct” metric information about these trees (diameter of the crown, height of the tree) to estimate the biomass of the forest and thus promote models such as that developed in our study. From this, it will be possible to establish carbon levels. In other words, some current criticisms directed at satellite images which are said to be lacking in information on forest understory are seen to be unjustified in the context of carbon levels. Instead of extrapolating vegetation radiance and making a connection between this and the biomass, it will probably be possible to directly estimate the latter from the measurement of largest trees of which there are at least between 5 and 10 individuals recognizable per hectare”.

It is a source of great satisfaction to Jean-François Bastin that part of his work is rooted in the immediate present and not in some hypothetical future. Since 2012, his data has been used by the WWF Congo (World Wide Fund for Nature) which, in collaboration with NASA and Dr. Sassan Saatchi, is currently being used to map all the forests of the Congo basin in the context of conservation programmes funded by WWF Germany. In addition, the article that appeared in Scientific Reports has enabled the researcher to have access to a much larger database, and to use this to test his results on a pan-tropical scale. In the month of October, he will be hired by the United Nations Food and Agriculture Organization (UN-FAO), he will apply himself to creating a standard methodology for monitoring forests in the context of carbon credits.