气候变化领域集成服务门户

Reforestation in the tropics mitigates the negative effects of climate change by sequestering carbon in biomass. However, tree growth is limited by nutrient availability in many tropical regions. A clear understanding of nutrient constraints and topography on growth of native timber species is thus essential to improve both the economic return on reforestation and the ecosystem services in tropical degraded lands. To address this, we use 7-year growth data from a 75-ha reforestation experiment in central Panama to test a modeling approach to predict growth of these species. The experiment includes five valuable timber species in 21 treatments, including monocultures and mixtures. We first fit a non-linear growth model as a function of tree age, then expand the former model parameters as a function of variables related to species mixture and micro-site soil conditions. Finally, we built a final model for each species to predict growth along three axes: nutrient availability, slope and species mixture. The models successfully identified how variation in growth was related to micro-site conditions and the species mixture. Although all species were long-lived pioneers, most were overall more sensitive to nutrient availability and between-trees interactions than to slope. However, the fastest growing species on average was more sensitive to slope than the other species and less sensitive to nutrient availability, showing better performance than the other species even under adverse conditions. Our models aid identification of species with the best growth potential to use in reforestation on infertile soils, leading to a better species selection according to site conditions.