Quantifying the effect of the sampling plot size on the estimation of structural indicators in old-growth forest stands

There is increasing awareness that structure-based indicators should be considered for assessing the biological value of late successional forests. In order to increase the unique habitat features critical for old-growth associated species, it is important to identify and rank candidate potential forest habitats on the basis of their distinctive structural features. Data on living and deadwood components for the identification of old-growth condition are usually acquired in sample forest stands through two sampling strategies: i) census performed in large monitoring sites; ii) network of small sampling units, on which inventory practices are usually based. Several authors argued that choosing between these sampling strategies might have substantial effects on the values of common indicators of old-growth condition. Our study aims at: i) assessing the difference between four common indicators measured in field plots with different sizes; ii) defining the sample size for the assessment of old-growth structural indicators. The study was carried out in six beech dominated forest stands located on the Apennines range in Central Italy. In each stand, living and deadwood components were surveyed and geocoded in 1-ha wide areas. Circular plots with radius ranging from 4 m up to 20 m were then considered in order to quantify the effect of sampling plot size on the estimation of four structural indicators: 1) number of living trees; 2) number of large trees (dbh ≥ 50 cm); 3) total deadwood volume; 4) number of large deadwood elements (snags, dead standing trees; lying dead trees, lying deadwood) with dbh (or average diameter for lying deadwood) ≥ 30 cm. We found that the size of the sampling plots should be at least 500 m2 in order to establish a reliable database for the assessment of the investigated indicators. The census approach should always be preferred to the sampling plot approach for old-growth forest stands smaller than 2-3 ha. The achieved results contribute to define assessment protocols for characterizing and ranking the naturalness of forest stands and for supporting the identification of potential old-growth stands based on standardized indicators.