Performance assessment of two plotless sampling methods for density estimation applied to some Alpine forests of northeastern Italy

Abstract

In this study, we tested two plotless sampling methods, the ordered distance method and point-centred quarter method, to estimate the tree density and basal area in some managed Alpine forests in northeastern Italy. We selected nine independent forest stands, classified according to the spatial distribution patterns of trees (cluster, random, regular). A plotless sampling survey was simulated within the selected stands and the tree density and basal area were estimated by applying both the ordered distance method and point-centred quarter method. We compared the estimates, in terms of accuracy and precision, between the two methods and against estimates obtained from a simulated survey based on a plot-based sampling method. The point-centred quarter method outperformed the ordered distance method in terms of both accuracy and precision, showing higher robustness towards the bias related to non-random spatial patterns. However, both the plotless methods we tested can provide unbiased accuracy of estimates which, in addition, do not differ from estimates of plot-based sampling. The satisfactory results are encouraging for further tests over other Italian Alpine as well as Apennine forests. If confirmed, the plotless sampling method, especially the point-centred quarter method, could represent an effective alternative whenever plot-based sampling is deemed redundant, or expensive.

Keywords

• Distance-based Density Estimator
• Ordered Distance Method
• Point-centred Quarter Method
• Accuracy
• Precision
• Conditional Inference Trees