Tree-related Microhabitats (TreMs) are of prime concern for biodiversity since they host thousands of taxa. TreMs are discrete habitat patches borne by trees and cover a wide range of lifespans, from days to decades to centuries. The taxa associated with such discrete and sometimes highly ephemeral habitat patches are likely to be sensitive to spatial distribution because they need to search for new habitats after the occupied one disappears. Although many studies have recently been dedicated to TreMs, only very few have investigated their spatial distribution. Focusing on European beech-dominated forests, we used a European TreM database with 12,362 trees and 296 plots (ranging from local (0.1–1 ha) to landscape scale (10,000 ha)) to assess TreM spatial distribution patterns in long unmanaged forests. Then, with a TreM sub-database with 6828 trees and 21 plots, we assessed whether and if so, how harvesting changes spatial patterns at the plot scale. In long unmanaged forests, most TreMs showed a regular pattern at the plot scale and an aggregated pattern at larger scales. Tree diameter was the most influential factor for spatial patterns in TreMs. Spatial patterns at the plot scale in harvested forests differed slightly from those observed in unmanaged forests. To favor TreM-dwelling taxa in harvested stands, our results suggest retaining habitat trees in a regular spatial pattern to mimic the natural pattern. However, some TreMs should be conserved in an aggregate pattern; we specify the spatial scale at which this should be done.

Tree-related microhabitats (TreMs) are crucial for biodiversity, supporting thousands of species. Most TreM research has focused on temperate forests, prompting us to evaluate the TreM typology in the boreal biome. We aimed to: (i) identify factors influencing TreM occurrence on living trees, (ii) assess redundancy and complementarity of TreM supply among Scots pine, birch, and aspen in Fennoscandia, (iii) compare TreM occurrence between Fennoscandian and Canadian boreal forests, and (iv) determine the minimum sample size for assessing TreM diversity in large-scale inventories. We inventoried 1515 trees in long unmanaged forests across Fennoscandia and compared results with a Canadian dataset. Findings show that Scots pine, birch, and aspen support distinct and complementary TreM assemblages, emphasizing the importance of mixed stands for biodiversity. Over larger geographic areas, local plot context ? including climate, disturbance history, and biotic/abiotic factors ? was the main driver of TreM occurrence. Unlike temperate forests, tree diameter was not a significant driver in boreal forests. The variability of TreM assemblages within the boreal biome underscores the need for large sample sizes to accurately assess TreM diversity. Sampling 1000 trees is sufficient to assess the seven TreM forms, while a larger sample is required to capture all fifteen TreM groups.