Estructura, dinámica y regeneración de los bosques mixtos de alcornoque (Quercus suber L.) y quejigo moruno (Quercus canariensis Willd.) del sur de la Península Ibéricauna aproximación multiescala

Abstract

This Thesis addresses the study of the structure, dynamics and regeneration of Q. suber (cork oak)- Q. canariensis (Algerian oak) forests, located in Southern Spain (Los Alcornocales Natural Park, CádizMálaga). The main objectives were: i) to identify the historical and environmental factors that have shaped forest composition and structure; and ii) to analyse differential regeneration patterns of both oak species along gradients of resource heterogeneity. The analyses were approached at several scales combining forest inventory data, cartographic information, filed work and modelling techniques. As an introductory study we first analysed the distribution and regeneration of five Quercus and five Pinus species along climatic, edaphic and topographic gradients at regional scale (Andalusia), based on the Spanish Second Forest Inventory data (chapter 2). The following analyses were focused on the mixed Q. suber- Q. canariensis forests. First, we quantified forest composition changes during the 20th century from a series of historical forest inventories, and studied current forest structure along management and environmental gradients at various spatial scales (chapter 3). Second, seed sowing experiments were carried out along natural gradients of resource availability (light and water) in order to study inter-specific differences in the early stages of the regeneration cycle of Q. suber and Q. canariensis. Models of acorn survival to removal by animals (chapter 3) and models of seedling establishment (chapter 4) were parameterised as function of vegetation cover (light), soil water content, soil compaction and seed size. Results showed that at regional scale Quercus species segregated along a gradient of water availability, whereas pines were associated with a gradient of temperature. Pine's regeneration was more abundant in forest areas without tree cover, while for Quercus species’ regeneration occurred primarily under forest canopy. The distribution patterns found for Q. ilex and Q. faginea suggest that both oak species reach the coldest and driest areas thanks to the facilitation effect of pine canopy on oak seedling establishment. Last-century forest management has shaped Q. suber- Q. canariensis forest structure and composition severely. A sharp increase in the density of Q. suber trees was found. This species was clearly favoured for cork production at the expense of Q. canariensis, particularly in more rainy areas, providing further evidence for humans as major drivers of oak forest composition across the Mediterranean. The impact of management is imprinted on the present-day forest structure; Q. suber dominates in most of the stands, whereas Q. canariensis is associated with moister habitats near streams. Given that Q. canariensis stands are currently protected as forest habitat, it would be reasonable to predict an upward trend locally, in those areas where forest management has modified forest composition. Q. suber and Q. canariensis regeneration is scarce and seems to be limited by both biotic and abiotic factors. Once seeds reach the ground, a high proportion is removed by mice. The probability of seed removal increased with plant cover for the two oak species. Inter-specific differences in acorn removal were higher in open areas and disappeared in closed microhabitats, especially during a non-mast year. Despite later seed-drop, Q. suber acorns were removed at a higher proportion probably due to their larger seed size on average. In the next regeneration stages, autumn-winter heavy rains caused over-abundant soil water levels, associated with the more open habitats. Soil waterlogging reduced germination and emergence and lengthened time to emergence, which in turn decreased seedling survival during the dry season. Species mortality peaks occurred in different life stages. There seems to be a conflict between demographic stages; microhabitats with a low removal risk for seeds (open) may exhibit less favourable conditions for subsequent seedling recruitment, and vice versa. Similarly, seed size effect in each stage was opposite; while smaller seeds showed higher probability to survive predation, seedlings from smaller seeds had overall decreased probabilities of germination and emergence. Results suggest that between-year variation in precipitation could favour Q. suber and Q. canariensis species coexistence. We speculate that under not such rainy winters (without waterlogging) Q. suber would benefit, since it showed higher germination and emergence rates in humid but well drained soils. Conversely, Q. suber would be negatively affected by heavy rains due to its greater delay in emergence, which would result in a lower recruitment for this species. On the other hand, late spring rains and particularly sporadic summer rains seems to favour Q. canariensis. As a whole, the decrease of precipitations and increase of temperature predicted for the climate change process, will negatively affect both oak species regeneration.