We examined 6 direct (i.e., based on tree ring counts) and 2 indirect (i.e., based on correlations between age and tree characteristics) methods to estimate the age of hollow trees. The errors associated with methods were compared by simulating rotten centres of different sizes (diameter: 5–82 cm) in tree ring cores from sound old oaks (Quercus robur) (diameter: 17–129 cm) collected from pastures in southeast Sweden. The lowest error (mean deviation: 15%) was obtained using tree ring data from the sampled hollow trees in combination with a function based on the growth pattern of sound trees. Indirect methods resulted in moderate errors (mean deviation was 23% and 26% for a function with bark crevice depth and with tree diameter, respectively). Because rotten centres often develop asymmetrically, we found it desirable to take a minimum of 2 cores from each hollow tree, from different cardinal directions. Trunks with a high probability of having a rotten centre could be identified before coring, as the width of the rotten centre increased with increasing bark crevice depth, with increasing size of any entrance hole, and with decreasing distance between the lowest entrance hole and the ground.

Nomenclature : Lid, 1985.