Predicting regeneration establishment in Norway spruce plantations using a multivariate multilevel model |
| |
Authors: | Jari Miina Timo Saksa |
| |
Affiliation: | (1) The Finnish Forest Research Institute, Joensuu Research Unit, P.O. Box 68, FI-80101 Joensuu, Finland;(2) The Finnish Forest Research Institute, Suonenjoki Research Unit, FI-77600 Suonenjoki, Finland |
| |
Abstract: | This study predicts the regeneration establishment on 3-year-old Norway spruce (Picea abies (L.) Karst.) plantations in southern Finland using regeneration survey data. Regeneration establishment was described by seven response variables: number of planted spruces, natural Scots pines (Pinus sylvestris L.), natural spruces, natural seed-origin birches (Betula pubescens Ehrh. and B. pendula Roth.) and other broadleaves (i.e. sprout-origin birches and other broadleaves than birch), as well as height of crop-tree spruce and dominant height of broadleaves. Due to the multivariate (several responses for each plot) and multilevel (plot, stand, municipality, forest centre) structure, regeneration establishment was modelled by fitting a multivariate multilevel model with explanatory variables such as temperature sum, site fertility, soil quality and method of site preparation. In the model, the numbers of tree seedlings were modelled using over-dispersed Poisson distributed equations, and the tree heights were modelled using normally distributed linear equations. The estimated fixed and random parameters of the equations were logical, and there was no serious bias in predicting the regeneration establishment in the independent test data set. This modelling approach can be used to predict the regeneration establishment stochastically by taking into account the large unexplained variation in regeneration models. |
| |
Keywords: | Generalized linear mixed models Picea abies Poisson model Regeneration survey data Simultaneous models |
本文献已被 SpringerLink 等数据库收录! |
|