Structurally complex canopies create heterogeneity in resource availability (Hypothesis 2)
We confirmed our second hypothesis that structurally complex canopies increase heterogeneity of light and soil resources at the forest floor. The structural complexity of the tree canopy influenced the distribution of light within forests in space and time. In this study, structural complexity increased with an increasing number of gaps in the canopy, resulting in an heterogenous diversity of moving sun flecks at the forest floor. In these sun flecks, the amount of light can be over two orders of magnitude higher than in the shade (Chazdon and Pearcy 1991). This can be beneficial for some species which are able to handle the sudden increase, or detrimental for those which are sensitive to full sunlight. As the sun is moving during the day, the spatial configuration of leaves, twigs and stems will affect how long a certain spot or certain plant will be irradiated. Furthermore, the different incidence angles of the sun during the year will influence the spatial alignment of the sun flecks. This also affects in which phenological stage a plant is standing in full light or shade. Hence, on a very small scale numerous species could be adapted to the many different light conditions present over space and time. For soil resources, we found that C:N ratio and potassium content increased with canopy complexity, particularly for SSCI and DBHsd. A higher canopy complexity causes a heterogeneous distribution of plant material in forest stands. In combination with heterogeneous light distribution on the forest floor and eventually heterogeneous water input, decomposition of organic matter is altered at the small scale level which leads to heterogeneity of the C:N ratio. In addition, DBHsd also increased light heterogeneity. As the DBH is strongly correlated with tree age, results imply that a stand with different age classes will increase light heterogeneity. The correlation of potassium with canopy complexity was very weak although it is possible there is a covariance with other variables.