The aim of the present study was to examine the micromorphology of the cork-oak (Quercus suber) acorn and to evaluate the efficiency of the pericarp as a barrier to water loss. When the desiccation curve obtained at 25 ± 1°C and 62% RH for physiologically mature acorns collected from the tree was compared with that corresponding to fully ripe acorns which had been shed, the latter showed a slightly lower rate of water loss. A marked reduction in the mean rate of water loss was recorded for intact acorns compared with those from which the pericarp had been removed. The external surface and transverse sections of the pericarp from both mature and fully ripe acorns were examined by scanning electron microscopy. Two zones presenting morphological and micromorphological differences were identified: an area including the point of attachment to the cupule and an apical zone covering the embryo. The microstructure of the pericarp showed an external thick cuticle and a single external palisade layer of closely packed cells with no intercellular spaces. Underlying this layer, there was a further parenchymatous layer of poorly differentiated, roughly isodiametric cells. The pericarp in the cupular zone consisted of only this undifferentiated layer between the two epidermal cell layers and contained vascular bundles with many xylem elements.

Dehesas, rangelands occupied by scattered oak trees and characterized by silvopastoral uses, cover about 3.1 million ha in south-western Spain. There is considerable debate about the long-term persistence of holm oak (Quercus ilex) populations in dehesas, since most stands are overaged and seedlings and saplings are sparse. The forest cycle has been disrupted in most dehesas. Regeneration has been inhibited since stands were opened for agriculture and grazing. Oak diameters from three land-use groups (young dehesa [YD], middle-aged dehesa [MD], and old dehesa [OD]) in Cáceres Province, Spain, were compared. These groups differed in the age of the land-use system, i.e. time since the original Mediterranean forest was cleared. The dehesa systems were established about 80 (YD), 150 (MD) and 500 (OD) years ago. An analysis of 66 holm oak cross sections revealed a close correlation (r2 = 91.2%) between tree rings and diameters, so that diameter seems to be a reliable indicator of tree age. Nested analysis of variance showed significant variation in diameters between the land-use groups. There is generally a positive relationship between tree age and the age of agrosilvopastoral use of the dehesas. Sparse holm oaks in the dehesas are primarily remnants from the first forest cycle. Local differences in growth conditions (for example soil quality and tree density) contribute further significant diameter variation on a between-plot level. Diameter structure of abandoned dehesas showed two peaks and a high proportion of trees in the smallest size class. This indicates that the forest degradation process is reversible. An effective regeneration policy should promote a rotating 20- to 30-year set-aside of dehesa parcels.