Reconstructing past rockfall activity with tree rings: Some methodological considerations

This paper reports on methodological difficulties arising when trees and tree-ring series are used to reconstruct rockfall activity. Unlike disturbances from other processes such as debris flows, floods or snow avalanches, scars caused by rockfall are more randomly distributed on trees and may occur at considerable height on the stem. Therefore, the selection and the number of trees as well as the determination of sampling heights and sample depth, all require special care in investigations of spatio-temporal variations in rockfall activity. The purpose of this paper is to assess the reliability and potential standardization of methods used to determine rockfall impact heights occurring in trees. The data come from three case studies located in prealpine and alpine environments of Switzerland. The paper also addresses questions about the relationship between ‘hidden’ (masked) and visible scars. Furthermore, we investigate the relationship between scar data derived from one single cross-section or increment cores and the total population of scars on a stem.Our results indicate that impact scars may be identified from almost ground level up to 9 m of height on the stem surface. The wound healing and the masking of scars, in contrast, greatly depend on the bark properties of the tree species, their annual increment rates and age, as well as on the predominant size class of rockfall fragments. As a result, as many as 90% of all scars were masked in thick-barked Larix decidua Mill., whereas the ‘hidden’ scars only counted for 16–25% in Abies alba Mill. or Fagus sylvatica L. The analysis of single cross-sections at a given height shows that, at best, 13–37% of the rockfall scars occurring on the entire tree can be detected. The data suggest that the number of events reconstructed at a given height can be considerably improved if other signs of growth disturbances such as reaction wood, abrupt growth reductions or, when present, rows of traumatic resin ducts are considered as well.