Summary• The benefits or costs of mycorrhizal symbiosis for the host plants are most frequently estimated in terms of positive or negative effects on plant growth and biomass. The effect of mycorrhizal symbiosis on evolutionarily important fitness traits, such as reproduction, and effects on the next generation are poorly known.• In a set of glasshouse experiments involving two plant generations we investigated the impact of mycorrhiza on fitness traits in perennial hand-pollinated Campanula rotundifolia .• Mycorrhiza reduced plant growth and flower production, but increased rootshoot ratio, and shoot phosphorus (P) concentration. Mycorrhiza had no effect on seed number per capsule, mean seed weight or germination rate, but increased seed P concentration. Seedlings from mycorrhizal plants had a higher relative growth rate (RGR) than seedlings from nonmycorrhizal plants. Self-pollinations yielded fewer seeds per capsule that had inferior germination than seeds from cross-pollinations.• Mycorrhizal infection was associated with a cost for the reproductive output of the C. rotundifolia host plant, whose progeny, however, were better than those from the nonmycorrhizal plants. Mycorrhiza also showed the potential to affect the plant mating system by increasing self-incompatibility and inbreeding depression expressed during seedling growth.

Abstract. The deciduous Vaccinium myrtillus and the evergreen Vaccinium vitis‐idaea were subjected to five removal treatments of understorey layers: control, removal of the moss layer, removal of the field layer, removal of both moss and field layers and removal of moss, field and humus layers. A second factor, sowing, was included to investigate sexual reproduction after disturbance. Density of new ramets and seedlings and growth of annual shoots were studied for the first two growing seasons, whereas cover was measured for five growing seasons after disturbance treatment. Initially, vegetative production of new ramets and species cover increased rapidly in all disturbed plots, except for the most severe treatment, in which production of new ramets was virtually absent throughout the study. Full recovery following removal of the field layer only or both field and moss layers was reached after four years for V. myrtillus. V. vitis‐idaea recovered more quickly, after one year (removal of field layer only) and four years (removal of field and moss layers). The relative growth of V. myrtillus and V. vitis‐idaea increased in the latter treatment in terms of production of annual shoots and length of annual shoots, respectively. Seedling density increased after sowing in the most severe treatment. The results underscore the importance of vegetative growth for recovery of these species at moderate‐level disturbances. The high rate of sexual reproduction in the most severe treatment implies that strong mechanical disturbance is needed to enhance the establishment of new genotypes in these species.

Clonal propagation and limited dispersal of seeds lead to genetic population structures in which most potential mates are relatives. If the species suffers from inbreeding depression or is self-incompatible, the number of seeds matured per fruit may be limited by the lack of outcrossing by unrelated pollen. We tested for distance-related genetic structure by hand-pollinating plants at increasing distances and measuring fruit set (berries/pollinated flowers) and seed number per fruit in natural populations of bilberry (Vaccinium myrtillus L.) and lingonberry (Vaccinium vitis-idaea L.). Bagging of the flower buds without natural (pollination by insects) or hand-pollination yielded almost no fruit or seed, suggesting that the species are obligately insect pollinated. Self-pollination led to a significant reduction in seed number per fruit and increased seed abortion compared with cross-pollination. Moreover, seed number increased with increasing distance between the pollen donor and pollen recipient. Flight distances of bumblebees were estimated for bilberry by monitoring between-flower and, for lingonberry, between-inflorescence flight distances. About 90% of all flights were at distances of less than 1 m. At this distance, we observed less than the maximum seed number per fruit in hand-pollinations in both study species. Consequently, clonal growth is likely to be an important factor that constrains fruit and seed number in these species.Key words: bumblebee forage distance, clonal, hand-pollination, inter-parent distance, reproductive success, Vaccinium.

Many zoophilous plants attract their pollinators by offering nectar as a reward. In gynodioecious plants (i.e. populations are composed of female and hermaphrodite individuals) nectar production has been repeatedly reported to be larger in hermaphrodite compared to female flowers even though nectar production across the different floral phases in dichogamous plants (i.e. plants with time separation of pollen dispersal and stigma receptivity) has rarely been examined. In this study, sugar production in nectar standing crop and secretion rate were investigated in Geranium sylvaticum, a gynodioecious plant species with protandry (i.e. with hermaphrodite flowers releasing their pollen before the stigma is receptive). We found that flowers from hermaphrodites produced more nectar than female flowers in terms of total nectar sugar content. In addition, differences in nectar production among floral phases were found in hermaphrodite flowers but not in female flowers. In hermaphrodite flowers, maximum sugar content coincided with pollen presentation and declined slightly towards the female phase, indicating nectar reabsorption, whereas in female flowers sugar content did not differ between the floral phases. These differences in floral reward are discussed in relation to visitation patterns by pollinators and seed production in this species.

In boreal spruce forests that rarely experience extensive disturbances, fine-scale vegetation gaps are important for succession dynamics and species diversity. We examined the community implications of fine-scale gap disturbances by selective removal of vegetation layers in a pristine boreal spruce forest in Northern Finland. The aim was to investigate how the speed of recovery depends on the type of disturbance and the species growth form. We also wanted to know if there appeared changes in species composition after disturbance. Five different treatments were applied in the study: Control, removal of the ground layer (bryophytes and lichens), removal of the understorey layer (dwarf shrubs, herbs and graminoids), removal of both the ground and understorey layers, and complete removal of the vegetation and humus layers above the mineral soil. The vegetation recovery was monitored in terms of cover and species numbers over a 5-year period. Understorey layer cover, composed mainly of clonal dwarf shrubs, recovered completely in 4 years in treatments where the humus layer remained intact, whereas ground layer cover did not reach the control level in plots from where bryophytes and lichens were removed. Recovery was faster in terms of species number than species cover. Bryophytes, graminoids and dominant dwarf shrubs appeared in all disturbed plots quickly after disturbance. Seedlings of trees appeared exclusively in disturbed plots. Graminoids dominated after the removal of humus layer. The results indicate that the regeneration of forest floor after small gap disturbance occurs mainly by re-establishment of the dominant species. Although destruction of the humus layer leaves a long-lasting scar to the forest floor, exposing of mineral soil may enhance the sexual reproduction of dominant species and the colonization of weaker competitors.