Forests of the world

Classification of the tropical forests of the world

We recently participated in a scientific article published in the prestigious journal “Proceedings of the National Academy of Science of the USA” ( The paper entitled ‘a phylogenetic classification of the world’s tropical forests’ was about the classification or relatedness of trees in the tropical forests of the world. For most of us, a forest is a sum of trees, like say, a plantation. In reality, forests are very complex and have different species composition in different parts of the world. Forests of Asia are different in tree species composition than the forests of South-America. Trees may belong to different genera and different families. For example, the mango tree, Mangifera indica, a native of the Indian subcontinent, belongs to the Anacardiaceae family and the eucalyptus (from Australia), belong to the Myrtaceae family. Looking at many forest plots and the species they harbor, it is possible to detect differences in forest species composition. Our paper shows that the world’s tropical forests can be divided into five major floristic regions (Indo-Pacific, Subtropical, African, American, and Dry forests) and not the traditional neo- versus paleotropical forest division (America vs Africa-Asia). Why should it be so?

Because of the movement of continents at the surface of the earth, a huge continent, Gondwana, split into south-America, Africa in the west and Australia, Antartica, India in the east. The timing of the split of continents tends to explain how closely related trees are globally. In general, the closer trees are in space, the closer they tend to be genetically. The split of Gondwana and not just the formation of the Atlantic Ocean explains vegetation patterns. This is an amazing finding that shows again how important the unique geology of the earth is, for vegetation.

To obtain these results, the main author, Dr. Ferry Slik of the University of Brunei, put together a large database contributed by approximately 150 scientists. The dataset originally included 439 locations containing 925,009 individual trees! Such large networks are more and more frequent in ecological research and help analyze information no individual or laboratory could gather otherwise. It is unfortunate that in spite of all the efforts to know better the biosphere, we continue its destruction at an ever accelerating rate. This may not be wise because the biosphere is our only habitable world.

Jean-Philippe Puyravaud