2026-05-05T19:33:55Zhttps://uvadoc.uva.es/oai/request

oai:uvadoc.uva.es:10324/505242021-12-03T13:02:56Zcom_10324_7492com_10324_952com_10324_894col_10324_7493

00925njm 22002777a 4500 dc Rodríguez Puerta, Francisco author Lizarralde Torre, Íñigo author Bravo Oviedo, Felipe author 2013 A system of additive equations was developed to predict whole-tree volume and the different components of Corsican pine. In this work, the nonlinear seemingly unrelated regression (NSUR) approach, which guarantees additivity in nonlinear equations, was evaluated. The effect of bark thickness on the accuracy of the results for all tree components was also assessed. Data for 351 trees, ranging in age from 10 to 72 years, were collected from 65 public and private sites. The volume estimates show average biases that range in absolute values from 2.19 to 31.02 dm3 for whole-tree, from 1.41 to 27.31 dm3 for wood, and from 1.05 to 16.52 dm3 for bark volume components. Errors in volume predictions were relatively small, representing less than 3% of the average observed wood volume and less than 6% of the average observed bark volume. This research showed that satisfactory predictions can be obtained from forcing additivity using NSUR approach with a minimal number of easily measurable tree variables, such as dbh and total height. FOR. SCI. 59(4):464–471. Forest Science, 2013, vol. 59, n. 4, p. 464-471 0015-749X https://uvadoc.uva.es/handle/10324/50524 10.5849/forsci.12-023 464 4 471 Forest Science 59 Bosques y silvicultura - España Pinos - España Pinos - Crecimiento Bosques - Gestión - España Additivity on nonlinear stem taper functions: A case for Corsican pine in northern Spain