Seed dispersal effectiveness (SDE) is one of the best parameters for estimating the benefit that a plant obtains from a dispersal agent. SDE includes quantitative and qualitative components. Previous avian studies of SDE estimation have been conducted mainly in temperate shrub lands or in tropical forests; data are limited for temperate forests, especially, the quality of seed-deposition sites. In this study, we estimated the SDE of birds for Wild Cherry Cerasus jamasakura and Giant Dogwood Cornus controversa fruits in a temperate Japanese forest. These trees prefer sunny sites so we focused on the light conditions at seed-deposition sites and categorized three microhabitat types: forest-gap, forest-edge, and forest-interior. We observed avian seed removal at fruiting trees, conducted a bird census to estimate microhabitat selectivity and collected bird-dispersed seeds. We then counted seedling emergence. Finally, we constructed generalized linear mixed models that combined bird census and dispersed seed data to identify influential seed disperser(s) for each tree species. Frugivorous behavior was recorded for eight avian species for C. jamasakura and 12 for C. controversa. Of these birds, one for the former, and two for the latter preferentially selected forest-gap or forest-edge trees during each fruiting period. The density of dispersed seeds and seedlings was higher in the forest-gap or forest-edge. Additionally, the frequency with which Japanese White-eye Zosterops japonicus visited was a significant variable for seed density in each microhabitat. The results suggested that microhabitat selectivity may affect the density of dispersed seeds in each microhabitat. Furthermore, a bird with a high quantitative SDE may not necessarily also have a high qualitative SDE. Thus, it is important to estimate qualitative and quantitative SDE by focusing especially on the seed-deposition site for each bird in order to evaluate the actual SDE in temperate forests.