Abstract: A series of previous taper equations with variable parameters were compared in this study. The model with better fitting results (smaller residual sum of squares and higher correlation index) for Larix gmelinii taper was selected in terms of model fitting statistics (residual sum of squares and correlation index), and five indices for describing the taper were proposed, according to the meanings of variable parameters in the model. Among the taper equations compared, the equation provided by Lee et al. had better fitting effect, and could be used to describe the tapers of L. gmelinii plantation. The five indices including taper rate of root, influence point, ranges of parabolic and paraconic, minimum of variable parameters, and relative height at minimum of variable parameters could be taken as the methods and tools for comparing the tapers. Only moderate stand density (487 plant·hm^-2) provided good quality of larch stem form, compared to both high (870 plant·hm^-2) and low (275 plant·hm^-2) stand density providing poor quality.

Key words: allometric, leafing intensity, leaf mass, Larix gmelinii plantation, taper function, variable-exponent, density, stem form, bamboo, altitude, stem mass.