Data for "Trait-based response of deadwood and tree-related microhabitats to decline in temperate lowland and montane forests"

Sampling design and case studies The study was conducted in two French regions, the Loire valley and the French Pyrenees, and one German region, the Bavarian mountains. In the Loire valley, we studied two lowland sites in oak-dominated (both Quercus petraea (Matt.) Liebl. and Quercus robur L.) forests, one in the Orleans State Forest (107-174 m a.s.l.) and one in the Vierzon State Forest (120-190 m a.s.l.). The main secondary species in these forests were hornbeam (Carpinus betulus L.) and Scots pine (Pinus sylvestris L.). In 2020, we selected nine plots to represent a decline gradient in each of these forests. While the Orleans Forest was healthy overall, the Vierzon Forest had undergone several decline events due to successive droughts aggravated by edaphic factors. In the Pyrenees, we studied two sites in montane forests dominated by silver fir (Abies alba Mill.), whose decline is mainly the result of successive droughts occurring since the 1980’s, and with Norway spruce (Picea abies (L.) H. Karst) and European beech (Fagus sylvatica L.) as secondary species. In 2017, we selected 43 plots: (i) 21 plots in the Aure Valley (854-1570 m a.s.l.) and (ii) 22 plots on the Sault Plateau (705-1557 m a.s.l.). The severe summer drought of 2003 had significant effects on tree mortality in oak and fir forests (Cours and others, 2022). Finally, we studied 19 plots of montane forest in the Bavarian Forest National Park, dominated by Norway spruce (Picea abies (L.) H. Karst) with European beech and silver fir as the main secondary species (Bässler and others, 2009). The dieback results from several cycles of windstorms followed by bark beetle (Ips typographus (L.)) outbreaks (Müller and others, 2010), the dominant drivers of forest dynamics in Norway spruce forests in temperate Europe (Zemlerová and others, 2023). This dieback phenomenon was more severe than either of the aforementioned drought-induced declines, and resulted in greater tree mortality (Cours and others, 2021). In the fir and oak forests in France, our plots were set up in managed forests, and the surrounding forest was also predominantly managed. On the other hand, in the German spruce forest, our plots were set up both within the core area of the Bavarian Forest National Park, and in the surrounding zone (BIOKLIM project), with little or no human intervention (Müller and others, 2010). Field measurements Plots were set up with a Bitterlich relascope with an opening angle corresponding to counting factor n° 1 (ratio 1/50), and mean plot area was about 0.3 ha. For each tree within the plot, we recorded its status (i.e. dead, living, snag, log), tree-species and diameter at breast height (DBH; minimum DBH recorded = 17.5 cm for living trees and logs, 7.5 cm for snags, 67.5 cm for very large trees). We took the proportion of dead trees in basal area (i.e. the ratio of the cumulative basal area of standing and lying dead trees to the basal area of all the trees in the plot), hereinafter referred to as “mortality rate”, as a proxy for the level of local stand decline. Note that this “mortality rate” does not reflect true overall mortality rate in managed oak forests, as foresters removed most valuable declining trees. We visually inventoried TreMs on living trees, logs and snags, and included the 47 types described by Larrieu et al. (2018). For each deadwood item (length > 1 m) in the plot, we measured its decay stage (from 1 = hard dead wood fully covered with bark to 4 = soft wood without bark), length, diameter at mid-length for logs and snags < 4 m long, and DBH for dead trees and snags > 4 m. Deadwood was classified in the following categories: ground-lying (logs and uprooted dead trees) vs standing (snags and standing dead trees); small and mid-size (less than 40 cm in diameter) vs large and very large (more than 40 cm in diameter); and fresh (decay class 1 and 2) vs decayed (decay stage 3 and 4). We calculated the total number of items per hectare by allocating a coefficient Nd related to diameter (d) to each item observed in the relascope sampling: (Nd = π 108 [ArcTan(1/50)/(π d)]2). We estimated TreM diversity and the number of deadwood types per plot. We compiled a list of eco-morphological traits for woody elements (i.e., life status (living, dead) and vertical position (downed, standing), decay stage and diameter) and for TreMs detected in the field (TreM nature, association with deadwood (saproxylic, epixylic, mould), type of bearing substrate (i.e., living tree, dead tree or snag, and log), position in the tree (i.e. base, trunk, crown), degree of wetness, life span or ontogenesis).