Light is a key resource for plant growth and is of particular importance in forest ecosystems, because of the strong vertical structure leading to successive light interception from canopy to forest floor. Tree species differ in the quantity and heterogeneity of light they transmit. We expect decreases in both the quantity and spatial heterogeneity of light transmittance in mixed stands relative to monocultures, due to complementarity effects and niche filling. 2. We tested the degree to which tree species identity and diversity affected, via differences in tree and shrub cover, the spatiotemporal variation in light availability before, during and after leaf expansion. Plots with different combinations of three tree species with contrasting light transmittance were selected to obtain a diversity gradient from monocultures to three species mixtures. Light transmittance to the forest floor was measured with hemispherical photography. 3. Increased tree diversity led to increased canopy packing and decreased spatial light heterogeneity at the forest floor in all of the time periods. During leaf expansion, light transmittance did differ between the different tree species and timing of leaf expansion might thus be an important source of variation in light regimes for understory plant species. Although light transmittance at the canopy level after leaf expansion was not measured directly, it most likely differed between tree species and decreased in mixtures due to canopy packing. A complementary shrub layer led, however, to similar light levels at the forest floor in all species combinations in our plots. 4. Synthesis We find that a complementary shrub layer exploits the higher light availability in particular tree species combinations. Resources at the forest floor are thus ultimately determined by the combined effect of the tree and shrub layer. Mixing species led to less heterogeneity in the amount of light, reducing abiotic niche variability.

light availability datafileThe research was conducted across 53 plots (30 m x 30 m) located in mature forests in the region of Ghent, Belgium (the ‘TREEWEB’ platform) Light transmittance was measured with hemispherical images before and after leaf expansion at four locations in each of the 53 plots. We calculated the gap light index (GLI), which is the total transmitted PAR to the forest floor as a percentage of total incident PAR above the canopy.light_availability_driad.csv