Predicting annual stem diameter increment of major tree species in mixed hill dipterocarp forest with consideration on tree and stand level effects

Predicting tree species growth for tropical forests is a challenging task for forest researchers due to an insufficient number of samples needed to run the classical fixed-effects regression method. the aim of the paper was to develop a predictive model for annual diameter increment of a multispecies hill dipterocarp forest stand. to address this issue,a linear mixed-effects model was applied to predict the diameter at breast height (dbh) annual increment of trees that treated species as a random effect. a random coefficient for the predictor variables,which included tree size and competition parameters was examined. basic models using dbh and natural logarithm of dbh or log(dbh) were tested by treating the species random effects at various combinations of the predictor variables. the basic model that was selected contained random coefficients for intercept terms,dbh and log(dbh). subsequent predictor variables depicting the effects of competition from surrounding trees,trees smaller or larger than the subject trees,and a structure component represented by maximum dbh,were tested. the final model contained the predictors dbh,log(dbh),stocking density,maximum dbh and competition from trees smaller than the subject tree. © forest research institute malaysia.