s:2507:"TI Genetic variation in growth of Guazuma crinita (Mart.) trees at an early age in the Peruvian Amazon 
AU Rochon C 
AU Margolis H A 
AU Weber J C 
AB Guazuma crinita (Mart.) is an important timber-tree species for agroforestry plantations in the Peruvian Amazon Basin. Trees are typically harvested at 6–10 years of age, but demand for wood is forcing farmers and loggers to harvest trees at increasingly younger ages. Previous research in farming communities led to the recommendation that local seed sources should be used for agroforestry plantations. Therefore, a provenance/progeny test of G. crinita was established in farming communities to determine if there was significant genetic variation within a representative watershed. The test includes 7 provenances and 200 families from the Aguaytía watershed, and is replicated in three planting zones (upper, middle and lower elevations) in the same watershed. Tree height and stem diameter were analyzed at 12 months across and within planting zones. In the analysis across planting zones, there was significant variation in height and diameter due to provenances, but in height only due to families within provenances. Means for provenances and families were relatively stable across zones, suggesting that the best provenance and families within that provenance could be used for agroforestry projects throughout the watershed. Variation due to provenances and families within provenances was most significant in the upper planting zone where trees grew most rapidly. Heritability of height and diameter was relatively low across zones, but again was notably higher in the planting zone where trees grew most rapidly. Results suggest that there is potential to improve yield of G. crinita by selecting faster-growing provenances and families within watersheds. However, selection would not be very efficient at 12 months since heritability was quite low at this stage of development (0–0.14 depending on planting zone). Heritability of growth traits, genetic correlations between growth and wood traits, and variation in mortality should be evaluated for at least 3 more years before deciding on the tree improvement strategy. To our knowledge, this is the first study to demonstrate significant variation due to provenances and families within a single small watershed in the Amazon. Results suggest that watershed-scale genetic studies should be conducted for other tropical hardwood species that are of interest to local communities. 
";