117 Copyright © ECOLOGICAL BULLETINS, 2010 Ecological Bulletins 53: 117–129, 2010 Tree mortality is a fundamental element of natural forest dynamics. Typically, annual overall tree mortality rates of canopy trees are in a range within 0.5 and 2% with only small variation between tree species (Runkle 1985, 2000, Wolf et al. 2004, Drobyshev et al. 2009). Insights into mortality dynamics often come from the studies of natu- ral disturbance regimes (Runkle 1985, 2000, Clark 1991, Esseen 1994, Wolf et al. 2004, Chapman et al. 2008). A number of studies have addressed variation in mortality rates as a function of tree characteristics and, particularly, tree size (Franklin et al. 1987, Mueller-Dombois 1987, Waring 1987, Dobbertin and Biging 1998). Small trees under closed canopies and during the early successional stages are often subject to high mortality due to self-thin- ning and limited amounts of light. A low mortality period tends to be associated with intermediate diameter classes, followed by an age- and size-related increase in mortality rate (Franklin et al. 1987). This general pattern, however, may vary as a function of tree life strategy in the commu- nity (Shimatani et al. 2008, Drobyshev et al. 2009). Mortality factors for canopy trees differ among ecosys- tems. In deciduous high canopy forests wind disturbance is often the dominant mortality factor and the main stem capacity to withstand breakage and uprooting are principal determinants of trees’ survival rates (Peterson and Pickett 1991, Ulanova 2000). To assess mortality levels, large da- Tree mortality in a semi-natural beech forest in SW Sweden Mauricio Fuentes, Mats Niklasson, Igor Drobyshev and Matts Karlsson To provide a better understanding of natural stand dynamics of Scandinavian broadleaved forests we studied pattern of tree mortality in a semi-natural beech Fagus sylvatica forest at Biskopstorp, SW Sweden. The primary goals were 1) to assess beech mortality rates of different tree condition classes (TCC) and transition probabilities between them, and 2) to suggest likely mortality drivers in the studied beech population. Mortality rates and transition probabilities were calculated between pairs of six TCC (live trees, live high stumps, dead trees, dead high stumps, stumps, and logs) by using data from two complete censuses of a 3.9 ha stand (n trees = 1189) over a ten year interval (1996–2006). The total beech mortality rate reached 2.0% yr –1 following two main pathways: 1) stem breakage leading to instantane- ous death (1.1% yr –1 ), and 2) stem breakage leading to partial crown removal, resulting in formation of a living high stump, HST L (10.5% yr –1 ) which was characterized by an increased mortality rate (11.5% yr –1 ). HST L made up ca 10% of the living trees in both censuses indicating the importance of this feature in the dynamics of the beech forest. Only two uproots were recorded (0.04% yr –1 ), despite of two hurricanes (in 1999 and 2005) that occurred during the study period. Mortality was significantly higher in trees above 60 cm DBH than in smaller trees (4.0 vs 0.7% yr –1 , respectively) and no differences were found among smaller DBH classes. Trees with fungal fruit bodies (FFB) had a significantly higher mortality than trees lacking FFB (7.6 vs 1.1% yr –1 , respectively). Our results highlight the role of the living high stump stage in the dynamics of the beech forest and suggest that the combination of wind and fungal infection of Fomes fomentarius are the main drivers of mortality in the studied stand. M. Fuentes and M. Niklasson, Swedish Univ. of Agricultural Sciences, Southern Swedish Forest Research Centre, P.O. Box 49, SE-230 53 Alnarp, Sweden. – I. Drobyshev (Igor.Drobyshev@ess.slu.se), Swedish Univ. of Agricultural Sciences, Southern Swedish Forest Research Centre, P.O. Box 49, SE-230 53 Alnarp, Sweden, and Univ. of Québec at Abitibi- Témiscamingue, Canada Research Chair in Ecology and Sustainable Forest Management, 445 boul de l’Université, Rouyn- Noranda, QC, Canada, J9X5E4. – M. Karlsson, Högestad and Christinehof Förvaltnings AB, Högestad, SE-271 97 Ystad, Sweden.