The present work proposes new boundaries for the current submediterranean territories of the Iberian Peninsula, defining them at the smallest scale attempted to date. The boundaries proposed are not sharp divisions but somewhat 'gradual', reflecting the transitional nature of the territories they encompass. Climate change predictions were used to estimate how the distribution of these submediterranean regions might change in the near future. The maps constructed are based on the distribution of marcescent Quercus species-trees that characterise the submediterranean plant landscape where they form the main forest communities. To determine their climatic range, the distribution of different types of Iberian oak forest was represented in 'climate diagrams' (ordination diagrams derived from principal components analysis), both in terms of individual species and groups of species based on leaf ecophysiological type, i.e. marcescent (Submediterranean), sclerophyllous (Mediterranean), semideciduous (Mediterranean) and deciduous (Eurosiberian). The climate range of each type of forest was determined, and the means of representative climate variables are analysed by one way ANOVA. The variables differentiating the forest groups were also examined by discriminant analysis. The range of the climate variables found to be associated with the majority of marcescent forests was used to determine the distribution of territories throughout the Peninsula with the same conditions (i.e. whether marcescent forests were present or not), thus providing a map of the Iberian submediterranean territories. Predictions of climate change were used to investigate possible climateinduced modifications in the boundaries of these territories in the near future. The patterns obtained show dramatic reductions in the extension of the Iberian submediterranean environment. Submediterranean conditions will probably disappear from the areas where they currently reign, and it seems unlikely that any new, large submediterranean areas will form by displacement towards higher altitudes. The outlook for the unique submediterranean vegetation of the Iberian Peninsula is gloomy.

We present a new stratigraphic (pollen and nonpollen microfossils and charcoal particle) sequence with five AMS dates, covering about the last 9000 yr, of a core collected from the Spanish northern meseta, one of the territories of the Iberian Peninsula for which little paleobotanical information is available. The results support the hypothesis of the permanence of the pine forests, in more or less pure masses or large timber stands, during the Holocene in some Iberian continental zones. The typical substitution in postglacial dynamics of heliophyllous species, such as pines or Cupressaceae, by broadleaf trees did not occur in this inland region. Presumably, factors linked to the substrate, in this case very deep sand dunes covering vast areas, may have contributed to the scarce local competitiveness of the broadleaf trees, which would account for the hegemony of pines in this region. Based on the dynamics of aquatic plants and nonpollen microfossils, an initial phase previous to 7500 14C yr B.P. of high levels in the lake can be identified by high percentages of hydrophytes. A progressive increase in helophytes and the gradual infilling of the lake over the last 5000 yr appear to indicate a phase of aridification similar to those established for the western Mediterranean. Study of charcoal particles, more abundant in the last 2500 yr, has given rise to certain hypotheses regarding the incidence of recurring fires of a local or regional nature.

Phenomena of hybridization can affect the ecology and evolution of the species involved in the process, as well as their communities. Although numerous papers focus upon the problem of taxonomy, few of these have attempted to study hybrid zones in relation to the analysis of their communities. On the Iberian Peninsula, hybridization phenomena among different oak species are frequent. It is, however, between Quercus faginea Lam. and Quercus pubescens Willd. where the most noteworthy hybridization phenomenon occurs. In this respect, we are familiar with the existence of different introgression levels but we are unaware of whether these hybrids are the transitory result of the interspecific genetic flow or whether these are maintained by means of extrinsic selection processes. Study of plant communities' flora and environment might shed light upon this issue. Comparison between plant communities dominated by one of the parental species and those dominated by individuals of hybrid origin might enable us to establish the presence or absence of an environment that is potentially selective in favour of the hybrids. Thus the possible existence of extrinsic selection. Furthermore, this information will help us to understand plant community distribution in an area that is difficult to interpret. To this purpose, we used multivariate ordination techniques (DCA and CCA) based upon a total of 395 floristic releves covering the whole range of the parental species on the Iberian Peninsula and upon climatic and edaphic variables for each of these releves. We also compared the groups obtained in relation to floristic similarity (Jaccard index), richness and diversity (ShannonWeaver index). Forests associated with Quercus pubescens are related to heavy summer precipitation, whereas Quercus faginea forests correspond to lower values of this variable and higher ones for continentality. Between both formations, there is a broad hybrid zone, with diffused borders that are related to an environmental gradient of Mediterranean influence. In this region, two types of forest communities were distinguished, which enabled us to divide the hybrid zones into two territories. Our results allowed us to locate the hybrid zone in an ecotone. The differentiation between habitats appears to indicate models of ecological selection. These models require, by definition, the presence of an environmental gradient between the parental zones. We are, however, aware of the need for future experiments in order to establish whether the hybrids are better adapted than the parental species. Only with availability of all this information can intrinsic selection be rejected.

No abstract

No abstract