Provenance research in Gmelina arborea Linn., Roxb. A summary of results from three decades of research and a discussion of how to use them

• Published in: The international forestry review 2002-01, Vol.4 (1), p.20-29
• Main Authors: LAURIDSEN, E. B., KJAER, E.D.
• Format: Article
• Language: English
• Subjects: Ecoregions; Forest management; Forestry research; Geographic regions; Plantation forestry; Plantations; Provenance; Trees; Tropical forestry; Wood density
• ISSN: 1465-5489; 2053-7778

More than 60 Gmelina arborea provenance trials were established in the mid-seventies as a joint effort between 20 countries, mainly outside the natural distribution areas of gmelina. Results from 27 of these trials are presented in this article focusing on differences between seed sources in growth rate, bole quality and wood density. The results allow identification of provenances that in general perform well on a number of sites. More detailed results are found on Danida Forest Seed Centre's web-site (www.dfsc.dk), which show performance of provenances in specific trials and include information on more characters. An interesting observation is that plants originating from plantations often perform very well compared to those originating from natural forests. This can be due to (i) a positive selection during thinning in the plantations, (ii) a result of lower rates of inbreeding in the plantations, or (iii) a positive response to adaptation to local conditions (because the plantings often represent local 'landraces'). The presented results cannot clarify the relative importance of these factors, but all three factors suggest that important gains can be obtained by tree improvement programmes in gmelina. The second potential factor (release of inbreeding) suggests that such programmes should incorporate a broad genetic base. They may include hybrids of we 11-performing provenances, and maintain genetic diversity through advanced generations. Plus de 60 essais de provenance de la Gmelina arborea furent établis au milieu des annés soixante-dix dans un effort commun de 20 pays, principalement hors des zones de distribution naturelle de la Gmelina. Les résultats de 27 de ces essais sont présentés dans cet article, en se concentrant sur les différences d'après la provenances des graines entre le régime de croissance, la qualité du tronc, et la densité du bois. Les résultats permettent l'identification de provenances qui croissent généralement bien sur un nombre de sites. Des résultats plus détaillés peuvent être obtenus sur le web-site du Danida Forest Seed Centre (www.dfsc.dk), montrant le résultat des provenances dans des tests spécifiques, et comprenant des informations sur d'autres aspects. Une observation intéressante est que les plantes originaires de plantations produisent des résultats souvent très bons comparés à celles originaires des forêts naturelles. Cela peut être dû à : 1) une sélection positive durant le clairsemage dans les plantations, 2) un