Difference methods based on six growth equations namely: Richards, Weibull, Korf, Logistic, Schumacher and Sloboda were adopted to build polymorphic dominant height and site index equations for Chinese fir plantations in Southern China. Data from stem analysis of 157 trees were used for the model construction. The performance of 15 equations including 10 kinds of difference equations was compared under different conditions. Effects on modeling precision caused by the variation of fitting data sets, site index, stands age and free parameter were analyzed and discussed. The results show that the attributes of inflection points of the biological growth equations had very important effects on their precision, while modeling dominant height. Difference equations had a better modeling precision for regional data sets than the prototypes of equations. The polymorphic dominant height equations, such as the two-parameter polymorphic form of Korf, Richards and Weibull and the three-parameter polymorphic form of Sloboda, showed higher precision. The two-parameter polymorphic form of Korf equation was selected to build a polymorphic site index equation for Chinese fir plantations.

Key words: Chinese fir plantations, difference equation, dominant height, polymorphic site index equations, precision.