Breakage or uprooting: How tree death type affects hillslope processes in old-growth temperate forests

Abstract

Tree breakage and uprooting are two possible scenarios of tree death that have differing effects on hillslope processes. In this study we aimed to (i) reveal the long-term structure of the biomechanical effects of trees (BETs) in relation to their radial growth and tree death types in four old-growth temperate forests in four different elevation settings with an altitudinal gradient of 152–1105ma.s.l., (ii) quantify affected areas and soil volumes associated with the studied BETs in reserves, and (iii) derive a general model of the role of BETs in hillslope processes in central European temperate forests.We analyzed the individual dynamics of circa 55,000 trees in an area of 161ha within four old-growth forests over 3–4 decades. Basal tree censuses established in all sites in the 1970s and repeated tree censuses in the 1990s and 2000s provided detailed information about the radial growth of each tree of DBH≥10cm as well as about types of tree death. We focused on the quantification of: (i) surviving still-living trees, (ii) new recruits, (iii) standing dead trees, (iv) uprooted trees, and (v) broken trees. Frequencies of phenomena were related to affected areas and volumes of soil using individual statistical models.The elevation contrasts were a significant factor in the structure of BETs. Differences between sites increased from frequencies of events through affected areas to volumes of soil associated with BETs. An average 2.7m3ha−1year−1 was associated with all BETs of the living and dying trees in lowlands, while there was an average of 7.8m3ha−1year−1 in the highest mountain site. Differences were caused mainly by the effects of dying trees. BETs associated with dead trees were 7–8 times larger in the mountains. Effects of dying trees and particularly treethrows represented about 70% of all BETs at both mountain sites, while it was 58% at the highland site and only 32% at the lowland site. Our results show a more significant role of BETs in hillslope processes including slope denudation in the mountains. We would expect a significant decrease of the biogeomorphic effect of trees in managed forests, but with a greater relative effect in mountains.