Community‐level seedling dynamics in Mediterranean forests: uncoupling between the canopy and the seedling layers

Abstract

AbstractQuestionsBased on comparisons between canopy and seedling layers, which woody species of a Mediterranean forest community, if any, are recruitment‐limited? What abiotic and biotic factors predominantly affect seedling density and survival for each species? How spatially consistent are the two demographic processes? What are the ecological implications of these findings for forest stand dynamics and species co‐existence?LocationMediterranean oak forests in southern Spain.MethodsWe present the results of a community‐level study conducted over 4 yr in three Mediterranean forest sites in order to determine species‐specific recruitment patterns. We analyse the most influential factors for seedling dynamics in the main co‐occurring woody plant species.ResultsWe found a strong uncoupling between the canopy and the seedling layer, irrespective of the structural characteristics of the forest site. Some of the most dominant species in the overstorey, such as Quercus suber, Arbutus unedo or Pistacia lenticus, were scarcely represented as seedlings in the understorey, suggesting problems of recruitment limitation. In contrast, other shrub species, such as Viburnum tinus or Phillyrea latifolia, showed large seedling densities with a high probability of survival despite their low frequencies as adults in the canopy. Species composition and abundance in the seedling layer were the result of the combined effect of different mechanisms involving both seed dispersal and the environmental filtering of seedling establishment. However, the factors that determined seedling spatial patterns were inconsistent with those that influenced seedling survival, probably as a consequence of different habitat associations through subsequent recruitment life stages.ConclusionsThe community‐wide approach enabled us to detect substantial differences among species in their spatial and temporal recruitment patterns, as well as in the relative importance of factors affecting seedling dynamics. The differential effect that spatial micro‐heterogeneity exerted for each of the studied species suggests the existence of multiple species‐specific regeneration niches, favouring species co‐existence and the maintenance of these diverse forest communities. However, the observed recruitment limitation for some dominant forest species will probably result in future shifts in species composition.