Comparison of forest age estimators using k-tree, fixed-radius, and variable-radius plot sampling

Quantifying the age characteristics of a forest can provide valuable information about the forest’s impact on the environment. For instance, the age of a forest can affect the ecosystem’s carbon exchange, soil enzyme activity, and biodiversity. In this paper, we investigate the use of different sampling methods to estimate the age characteristics of three simulated ponderosa pine (Pinus ponderosa Dougl. ex P. Lawson & C. Lawson) forests having different spatial and age patterns. This includes estimating the mean tree age and the age-class distribution of the trees in the forest. The trees in the sample are selected using k-tree sampling, fixed-radius plot sampling, or variable-radius plot sampling, and we compare the properties of the resulting estimators via design-based and model-based approaches. Analyses of the different sampling methods applied to the three forests suggest that the estimator associated with k-tree sampling, with the addition of a few extra trees per plot, is feasible for forests having a spatially mosaic or random spatial pattern. The estimator associated with fixed-radius plot sampling performed well for the forest having a clustered spatial pattern.