The forest layers of the canopy towards the ground are surfaces of exchange which condition an intraforestal environment different from the general climate.
The effects are mediated through light-sharing among different forest strata. The following paper describes the light conditions at the forest floor and the resulting regeneration patterns with respect to the structural dynamics of a natural forest:
Understorey light conditions and regeneration with respect to the structural dynamics of a near-natural temperate deciduous forest in Denmark, Forest Ecology and Management 106 (1998) 83–95, DOI: 10.1016/S0378-1127(97)00299-5
There are many issues that detail the interactions between forest storeys and forest regeneration. In a relatively stable, permanent forest environment, the relationships soil-plant-atmosphere are vital to the survival and growth of new specimens. Therefore, the needs of light and 'buffering' microclimate surfaces are one of the main targets to consider in the analysis of the composition and dynamics of forest structure. I would suggest focusing on the effects of climate variability (under changing climate conditions) on the availability (number, frequency, size) of microclimatic conditions suitable for natural regeneration during severe climate conditions. This may add very relevant information about the future survival OR changes in the composition of the forest.
Plant Species and the forests have natural way of controlling regeneration. for example, single dominat species like confers or teak, Eucalyptus,Acacia etc. have some allelopathic effects on forest floor, that prohibits other species to regenerate and grow. The other climax broad leaved forest with dense canopy have adaptation to develop seedlings in low light environment and light demanding species may find it difficult to grow. So, vegetation strata, many a times controls the species and the regeneration pattern.
In my opinion, an ecological approach to silviculture should avoid to rely much on generalizations. The response of seed and especially seedlings are still non completely known and they are very sensitive non only to microclimate and litter/humus type and quality but also to the influence and effects of the ground flora. Although theoretical knowledge is a necessary step, local cases/examples of the real behaviors of species under regeneration may be a very helpful tool to achieve successful forest management. Never forget that many plant species communicate and can transfer substances and information through the root network, and testing with preliminary trials is frequently a pragmatical and functional test to do.