It would be fair to say that knowledge of the forest types of lowland tropical Asia is greater than that of the neotropics, but not as complete as that of Africa, where the integrative role of A.E.T.F.A.T. (1959) has set an example that no other continent has so far succeeded in emulating. This collaboration resulted in continental vegetation maps of unprecedented uniformity and precision. Two pioneering efforts do exist in Asia though, which provide a basis at least for reviewing what we know, and drawing up a schematic classification as a basis for discussion and refinement. These are the landmark and remarkably detailed preliminary survey of the Forest Types of India and Burma (Champion, 1935), and its revision for modern India alone with S. K. Seth (1968); and the magnificent mapping program of the Institut de la Carte lnternationale du Tapis Vegetal, the University of Toulouse, and the French Institute at Pondicherry. Both rest on a combination of a wealth of personal experience on the ground, and a rather comprehensive though inevitably regionally patchy literature, equally essential for success. The comprehensive forest working plans which have been and remain a tradition of India, and formerly Burma, were particularly important sources of ground information.
Champion & Seth’s classification (Table 1) gives primary emphasis to physiognomy including leaf phenology, and structure including life forms and morphological characters, with less emphasis on species composition and dynamics. This was done explicitly in recognition of the influence of historical biogeography on the former, and of the confounding effects, and often local specificity of successional vegetation, widespread though it was even then. Their approach indeed provided a regional frame. But the downplaying of dynamic aspects led to a static rigidity which has made it inadequate in an era of actively managed, not to say degraded forests. The subordination of floristic information, and even of dominant species, to structure and physiognomy has obscured major regional differentiations which are related to geology and soil, rather than to climate and climate-correlated factors such as fire.
The Pondicherry group have consciously attempted to address both these shortcomings. Their method depends, first, on a classification of climate on criteria found, by empirical experience in many regions of the world, to be the most correlated with vegetation: Mean temperature of the coldest month, annual rainfall, and length and number of dry seasons. Second, it depends on comprehensive ground truth, extrapolated by means of aerial photographs, which has permitted use of dominant species as a major basis for their vegetation classification. In effect, soils-correlated variation on geographical scale is thereby incorporated, though the necessary soils survey data to define the interrelationship between vegetation and soil often does not exist. Comprehensive ground truth has also provided the means to more completely understand the dynamic relationships between different vegetations, and to more rigorously identify seres. The Pondicherry group have now extended their work from India including the Himalayas, and Cambodia to Indonesia, where they have published ‘bioclimate’ maps for Sumatra and Borneo, and a vegetation map of Sumatra.
View full abstract