Light regimes beneath closed canopies and tree-fall gaps are compared for five temperate and tropical forests using fish-eye photography of intact forest canopies and a model for calculating light penetration through idealized gaps. Beneath intact canopies, analyses of canopy photographs indicate that sunflecks potentially contribute 37–68% of seasonal total photosynthetically active radiation. In all of the forests, potential sunfleck duration is brief (4–6 min), but the frequency distributions of potential sunfleck duration vary because of differences in canopy geometry and recent disturbance history. Analysis of the photographs reveals that incidence angles for photosynthetically active radiation beneath closed canopies are not generally vertical for any of the forests, but there was considerable variation both among and within sites in the contribution of overhead versus low-angle lighting. Calculations of light penetration through idealized single-tree gaps in old growth Douglas-fir – hemlock forests indicate that such gaps have little effect on understory light regimes because of the high ratio of canopy height to gap diameter. However, single-tree gaps in the other four forest types produce significant overall increases in understory light levels. There is also significant spatial variation in seasonal total radiation in and around single-tree gaps. Our results demonstrate that there can be significant penetration of light into the understory adjacent to a gap, particularly at high latitudes. As gap size increases, both the mean and the range of light levels within the gap increases, but even in large gaps (ca. 1000 m²) the potential duration of direct sunlight is generally brief (<4 h). The major differences in gap light regimes of the five forests are largely a function of canopy height and latitude. The effects of latitude should also result in differences in gap light regimes across the geographic range of individual forest types.

We have analyzed the light transmission characteristics of the nine deciduous and coniferous species that dominate the transition oak–northern hardwood forests of southern New England. Maximum likelihood techniques were used to estimate species-specific light extinction coefficients, using fish-eye photography combined with data on the locations and geometry of trees in the neighborhood around each photo point. Quantum sensors were also used to quantify interspecific variation in the importance of sunflecks and beam enrichment. Variation in light extinction was closely correlated with shade tolerance and successional status of the species. The most shade-tolerant species (Fagusgrandifolia Ehrh. and Tsugacanadensis (L.) Carr.) cast the deepest shade (5% full sun). These differences were more closely related to differences in crown depth than to differences in light extinction per unit depth of crown. Sunflecks contributed relatively little radiation beneath late successional species (<10% of total understory photosynthetically active radiation), but represented a major fraction (40–50%) of radiation beneath less shade-tolerant species. Using growth and mortality functions for the same species developed in a related study, our results indicate that saplings of all of the species have high survivorship in the shade cast by conspecific adults. However, only the three most shade-tolerant species have low rates of sapling mortality under the low light levels characteristic of stands dominated by late successional species. Our results are consistent with previously reported models, which propose that secondary succession is driven by interspecific differences in resource uptake and tolerance.