Regeneration of a Mediterranean oak:: A whole-cycle approach

Using the holin oak (Quercus ilex) as a model system, we quantified the losses in the potential number of recruits from the flower to the sapling stage caused by abiotic factors, pre- and post-dispersal seed predators, and invertebrate and ungulate herbivores in one well-preserved forest plot (F) and in one dehesa (savanna-like) plot (D). Tree fecundity was an order of magnitude higher in D than in F. Predispersal (flower plus fruit) losses due to abiotic factors were much larger (71% in F and 90% in D) than those originated by biotic factors (29% in F and 10% in D). Post-dispersal predation of acorns under trees led to crop depletion in F, while predators did not deplete the acorn crop despite a much higher predation level in D. Surface acorns were invariably eaten, while most buried acorns survived to germination in both plots. Ninety-five percent of seedlings emerged under oaks in D and only 20% in F, the density of newly emerged seedlings being ten times greater in F. Emergence from experimentally sown acorns was very high on average (73%), while seedling survival to the 2-y sapling stage was primarily affected by water stress in the first summer, especially in the dehesa stand. A 75-fold difference in recruitment rates between populations was found (0.00150 in F and 0.00002 in D). This whole-cycle disparity was the result of differences in the conversion rate from viable fruits to newly emerged seedlings (two orders of magnitude larger in F). The inability to direct acorns to safe (shaded) sites by means of efficient dispersers appears to limit recruitment in open dehesa stands. This finding could explain the general lack of natural regeneration of dehesas as compared to the forests from which they developed.