A tentative composite climatic index to predict and quantify the effect of climate on natural rubber yield potential

Most of natural rubber (NR) comes from the cultivation of Hevea brasiliensis in tropical countries. However, mainly due to land use changes (urbanization, substitution...),new rubber plantation developments are often moving to climatically less favorable areas (usually called "marginal" areas or "non-traditional" areas). These new planting areas are often characterized by one or more climatic factors considered as less favorable to rubber tree cultivation. Howevern it is still a challenge to estimate exante and accurately the effect of such climatic stresses on the rubber yield potential. For this purpose, we propose a tentative multilinear regresstion model, using climatic parameters commonly available from local weather stations (total annual rainfall (mm), average thermal amplitude), in order to calculate a composite climatic marginality index for rubber tree cultivation (CCMIndex). Such CCMIndex could be used exante to provide an early evaluation of the expectable loss in rubber yield potential due to climatic marginality in any nex rubber planting area. It should also allow to discriminate among climatic types (climatic typology with 4 main types: warm/humid, warm/dry, cold/humid and cold/dry) and to quantify their respective climatic stress intensity. The CCM index could be used to adapt latex harvesting systems recommendations per climatic typeafter the type and the intensity of the climatic marginality have been descripted and quantified. Another index, based on dry months distribution along the year, can provide complementary information to describe the risk of occurrence of secondary leaf fall due to leaf diseases (SLF Risk Index). Combining the CCM index and the SLF Risk index could likely provide an accurate and simple estimate of the overall effect to climate on NR yield potential in any NR production area. Moreover, in a general context of climatic change, these two indexes might be used simultaneously to simulate the possible effect of future variations in temperature and rainfall intensity or pattern resulting from climatic changes on the future rubber yield potential of any current or future NR production area, at local, regional, national or global scales. This communication presents the current status of the model, which may be regularly upgraded depending on data accumulation improving the model parameterization.