Background: The lack of regeneration of cork oak (Quercus suber) is putting at risk the conservation of the montado. Biotic and abiotic factors as well as management are thought to constrain regeneration. Aims: Identify limitations to regeneration during the early stages and relate them to management practices. Methods: Experimental burying of acorns was carried out at three distinctly managed montados. Cache predation, seedling emergence, survival and resprouting were monitored during the first year. Regeneration was evaluated by testing the effect of biotic and abiotic factors in each stage. Results: Regeneration was determined by temperature and relative humidity. Early emerged seedlings showed higher drought tolerance and resprouting ability. Bigger acorns and higher plant cover significantly assisted emergence, survival and resprouting. Mortality causes were drought (38%) and mechanical ploughing (33%), while cache predation and browsing on seedlings were low (1.7%, when excluding tag effect on cache predation, and 0.6%, respectively). Seedling emergence and survival of drought were significantly lower in the most disturbed site than in the other two sites. Conclusions: Abiotic factors restrain regeneration while acorn size and plant cover are key factors determining seedling emergence and survival. The lack of regeneration may be overcome in the short term by preserving the shrub layer. Moreover, reforestation by direct sowing should be also considered.

a b s t r a c tThe future of the montado, a human shaped agro-forestry ecosystem of South Western Europe, is questioned due to the observed lack of cork oak health and low natural regeneration. We developed a System Dynamics Modelling approach to predict the long-term sustainability of this agro-forest, by recreating cork-oak population dynamics, management practices and the main environmental and biological constrains associated with this ecosystem. Our results indicate that the leading limitations to cork oak regeneration in montado ecosystems result from the intensity and interaction of land management practices, namely livestock and the use of heavy machinery. The main conclusions indicate that limiting the quantity of livestock up to 0.40 LU.ha −1 , and considering soil ploughing with a minimum periodicity of 5 years, are crucial to maintaining sustainable cork oak populations. This study represents a first step to support strategic options for cork oak montado management by providing projections of long-term population trends under realistic social-ecological change scenarios.

The invasive Clethra arborea has a dualrole in the diet of the Azores bullfinch, a critically endangered bird species endemic to the island of São Miguel (Azores, Portugal). This is a crucial winter food resource but it lowers the availability of native laurel forest species that compose most of the bird's diet throughout the year. The removal of this and other invasive alien species is part of current laurel forest habitat restoration programmes, disregarding the impact on the Azores bullfinch population. In order to evaluate the first responses of the Azores bullfinch to habitat restoration, we studied bird diet, foraging behaviour, food availability and habitat occupancy in managed (without C. arborea) and control areas. Significant increases in the availability of native food resources in managed areas were noticeable in the diet, particularly the intake of Ilex perado ssp. azorica and Prunus lusitanica ssp. azorica flower buds. In most of the studied months birds heavily used and foraged in managed over control areas. The one exception was in December, when a resource-gap occurred in managed areas, which may be overcome in the short-term due to re-establishment of native plants following removal of invasive aliens.

Few vertebrates are known to consume ferns regularly. Several species of mammals consume leaves to some extent but the consumption of fern spores is much rare. In Galicia (Northwest Spain) we studied the seasonal variation in the consumption of Culcita macrocarpa fertile leaves (i.e. with spores) in two populations (Capelada and Eume), assessed whether consumption rate increased with fern population size, and evaluated whether the consumer was a spore predator or a spore disperser. Consumption began in December and finished by mid February, and occurred before spore release, which happened later in Capelada than in Eume, probably influenced by differences in altitude. The consumer was identified as Apodemus sylvaticus by DNA analysis of its droppings and by capture of live animals. Throughout Galicia there was a significant increase in fern consumption rate as the population size of C. macrocarpa increased. Germination tests from droppings were carried out in 14 dishes but only in two dishes 1% and 0.3% of the spores germinated. Our results suggest that woodmouse can disperse spores of C. macrocarpa, although most of the spores were digested.