In a field study, we measured saplings of beech, ash and maple growing in a fairly even-aged mixed-species thicket established by natural regeneration beneath a patchy shelterwood canopy with 3–60% of above canopy radiation reaching the saplings. Under low light conditions, maple and ash showed a slight lead in recent annual length increment compared with beech. With increasing light, ash and maple constantly gained superiority in length increment, whereas beech approached an asymptotic value above 35% light. A suite of architectural and leaf morphological attributes indicated a more pronounced ability of beech to adapt to shade than ash and maple. Beech displayed its leaves along the entire tree height (with a concentration in the middle crown), yielding a higher live crown ratio than ash and maple. It allocated biomass preferentially to radial growth which resulted in low height to diameter ratios, and expressed marked plagiotropic growth in shade indicating a horizontal light-foraging strategy. In addition, beech exhibited the highest specific leaf area, a greater total leaf area per unit tree height, a slightly greater leaf area index, and a greater plasticity to light in total leaf area. Ash and maple presented a “gap species” growth strategy, characterized by a marked and constant response in growth rates to increasing light and an inability to strongly reduce their growth rates in deep shade. In shade, they showed some plasticity in displaying most of their leaf area at the top of the crown to minimize self-shading and to enhance light interception. Through this, particularly, maple developed an “umbrella” like crown. These species-specific responses may be used for controlling the development of mixed-species regeneration in shelterwood systems.

We compared shade tolerance of maple, ash and beech in the sapling stage from two sites with rich soils differing in water supply, growing in dense thickets underneath a beech shelterwood of varying canopy densities. Shade tolerance was described by two components: mortality in shade and height growth in high light. At low light, beech showed the least mortality, maple the highest and ash in between on both sites. The decline with increasing light was steepest in beech and more gradual with ash and maple. At ∼15 per cent above canopy light, all three species approached zero mortality. Beech as the most shade-tolerant species had the highest survival rate under low light and the least length growth rate under high light (>17 per cent). Ash had a lower survival rate at low light than beech and a highest growth rate at high light. Maple showed a bit weaker trade-off with the lowest survival rate but a growth rate inferior to ash. On the better water-supplied site, height growth was significantly superior in all three species only under high light. On the basis of these results, silvicultural conclusions are drawn with respect to appropriate light levels and cutting types.