The Polish Provenances of European Larch Overperform the Expected Growth Dynamics Indicated by the Sigmoid Model

Abstract

This article attempts an unusual interpretation of the observations characteristic of experiments that compare different tree species’ provenances. The focus falls on larch (Larix decidua subsp. polonica). The data came from the experiment established in 1967 at the Siemianice forest experimental station, where 21 Polish larch provenances were compared. The study’s main objective was to compare the basic estimates of growth dynamics, the maximum growth rate and acceleration, and the point in time when these values were achieved. A four-parameter sigmoid growth function was used to model the average stand basal area increments and its first and second derivatives to calculate the indicators of the growth dynamic. The models explained 98% of the observed variations resulting from the 21 inventories. Only one growth parameter showed a statistically significant difference among the compared provenances. The Góra Chełmowa provenance achieved the highest value of the maximum growth acceleration, but it was statistically significantly different only from the three underperformed provenances. However, when the average values for all the experimental plots (n = 86) were compared with those of the deterministic model (the stand volume and yield tables), the maximum growth rate and acceleration values were higher for the former. We discuss the potential factors responsible for this overperformance and point out the potential risks that arise from growth and quality metrics only when deciding on the best-performing provenances. The sigmoid growth model employed in this study might be an excellent alternative for comparing the growth dynamics among different stands or replications in experimental studies. Considering only the early results, the sigmoid growth model proves its limitations, and the conclusions reached should be treated with caution.