Résumé
The forest tree species Hevea brasiliensis is extensively planted in the humid tropics to meet the increasing demand for natural rubber. Huge quantities of planting stock are therefore needed. The seed option remains the easiest and cheapest way to establish plantations of rubber trees but those show a great variability for vigor and also for latex yield. The rationale of producing clones for overcoming this variability was already obvious in the early 1910's but due to the difficulties encountered at that time for rooting shoots, grafting was used as an alternative cloning method. The striking increase in yield noticed from the graft-derived clonal plantations warranted their large scale development. Eventually, the budded clones by virtue of their much higher and uniform productivity supplanted the seedlings in most industrial plantations. However, grafting is also associated with drawbacks and for decades efforts aiming at mass producing selected rubber tree clones on their own roots by rooted cuttings have been pursued. However, this approach was progressively abandoned due to disappointing rooting results and, from the 70's onwards, priority has been given to in vitro methods which were booming during this period. But despite 40 years of heavy investments, industrial H. brasiliensis clones could still not be mass micropropagated in vitro efficiently enough to meet the requirements of large scale production. The situation may change radically soon, however, due to the development of new nursery techniques adapted to the mass clonal production by rooted cuttings of any H. brasiliensis selected genotype. Efforts to improve the techniques as well as the establishment of new field trials are underway in order to determine if self-rooted rubber tree clones are more productive than grafted ones. This old issue is becoming of overriding importance considering the increasing pressure on land availability reducing thereby the prospects for expanding rubber tree plantations.