Eastern cottonwood clonal mixing study: predicted diameter distributions

Abstract

A parameter recovery procedure for the Weibull distribution function was modified to incorporate monocultures and mixtures of eastern cottonwood (Populusdeltoides Bartr.) clones planted in Mississippi and Kentucky. Components of the system included functions to predict stand-level basal area and four percentiles (0th, 25th, 50th, and 95th) of the cumulative diameter distribution. Basal area was predicted as a function of surviving number of trees, dominant height, age, planting location, and the proportion of each clone planted. Clonal proportions, which accounted for 3.6% of the variation in observed basal area, were more important than differences in planting locations, which accounted for 3.0% of the variation. Interactions between clones in mixtures were not significant (p = 0.5676), but some cases of both over- and under-compensation appeared to be developing. Percentiles of the cumulative diameter distribution were predicted as functions of quadratic mean diameter, and therefore included indirect effects of both genetic and planting site differences. Only the minimum diameter (D0) was directly affected by proportions of clones planted. Most of the monocultures and mixtures of clones had smaller minimum diameters than expected for a given value of quadratic mean diameter. The predicted quadratic mean diameter and percentiles were used to recover parameters of the Weibull distribution such that the predicted diameter distribution has the same quadratic mean diameter as obtained from the stand basal area model. The predicted distributions indicated that a common stand-level model was not sufficient for accounting for variations in diameter distributions of eastern cottonwood clones. As a result of the differences in diameter distributions, monocultures and mixtures of the Texas clones appeared to have less volume and greater stand variance than the Mississippi clones.