Revista Chapingo. Serie ciencias forestales y del ambiente
versão impressa ISSN 0186-3231
BUENO LOPEZ, Santiago W. e BEVILACQUA, Eddie. Developing a diameter-distribution prediction system for Pinus occidentalis, Sw. in la sierra, Dominican Republic. Rev. Chapingo [online]. 2011, vol.17, n.1, pp.115-132. ISSN 0186-3231. http://dx.doi.org/10.5154/r.rchscfa.2010.05.029.
Information about the diameter distribution of a forest stand is an essential component for deriving a variety of stand attributes, understanding stand dynamics under varying ecological conditions and planning future forest prescriptions. Diameter distributions from multiple evenaged stands of Pinus occidentalis from three different ecological zones within La Sierra region in the Dominican Republic were modeled using a three-parameter Weibull probability prediction method using either maximum likelihood or modified method of moments, and two bases on parameter recovery method using either percentile-based or cumulative distribution function regression. Stand characteristics such as age, total tree height, trees per hectare, site quality and basal area were employed as predictor variables. Methods were evaluated based on goodness-of-fit of the distributions and RMSE of stand yield. In these data, the parameter recovery method based on percentile based distribution proved to be considerably more efficient in describing the diameter distribution.
Palavras-chave : Pinus occidentalis; diameter distribution; Weibull pdf; parameter prediction; parameter recovery; mixed models.