Differences between the responses of European larch (Larix decidua Mill.) and Norway spruce (Picea abies (L.) Karst) to landslide activity based on dendrogeomorphic and dendrometric data

Abstract

Investigations of mass movement hazards need to be supported by accurate data regarding past frequencies of landslide reactivations. Dating landslide events using tree rings is a reliable approach for vegetated slopes in temperate climates. Despite the enhanced use of dendrogeomorphic research during recent decades, little is known about the specific responses of Larix spp. to landslide movements. We investigated the active section (~0.6 ha) of a complex landslide in the Vsetínské vrchy Mts. (Outer Western Carpathians) that is overgrown by a forest stand composed of European larch (Larix decidua Mill.) and Norway spruce (Picea abies (L.) Karst.). Using the dendrometric and dendrogeomorphic data identified from 74 trees (33 L. decidua and 41 P. abies), including the number and types of growth disturbances (GDs) and compiled event chronologies, we compared the responses of both tree species to the same landslide reactivations.We observed the same mean tree stem inclination for both tree species, whereas the inclination azimuths differed. L. decidua generally showed a higher number of stem bends compared to P. abies. Similar proportions of subjectively identified GDs (compression wood, CW, and abrupt growth suppression, GS) and mathematically identified GDs (tree-ring eccentricities) were observed in both tree species. However, the total number of GD.tree−1 was slightly higher in L. decidua (4.1) than in P. abies (3.2). A higher abundance of GS was identified in L. decidua at the expense of CW compared to P. abies. The event chronology based on the mathematical approach revealed slightly more landslide signals than for the subjective approach, mainly in L. decidua. However, the largest known landslide reactivation (in 1997) was recorded in all cases in both tree species. Based on the CW analysis, it was shown that larger stem inclinations are associated with longer persistence of CW in P. abies (rather than L. decidua), resulting in the possible masking of younger landslide events. By contrast, P. abies compensates for that limitation by having a generally shorter mean duration of CW than L. decidua. Overall, it seems that the complexity of landslide movements and root system architecture influenced the number and character of GDs in both tree species. In addition, we did not find any relationships between the measured dendrometric parameters and number of GDs to better assess the most suitable trees for dendrogeomorphic dating prior to sampling. Therefore, we recommend sampling both tree species when they overgrow the same landslide body so that the final event chronology will be more robust and possible limitations of one tree species can be eliminated by the benefits of the second.