Interaction between Stipa tenacissima and Pinus halepensis:: consequences for reforestation and the dynamics of grass steppes in semi-arid Mediterranean areas

Alpha or esparto grass (Stipa tenacissima) is a perennial tussock grass that coexists with Aleppo pines (Pinus halepensis) in semi-arid areas of southeast Spain and northern Africa where the pine is the tree most widely used in reforestation projects. Several authors have proposed that the improvement of soil characteristics by S. tenacissima in semi-arid slopes might be brought about in restoration programs by the introduction of shrubs and trees in these areas. In this paper we analyse the role of S. tenacissima as a nurse plant of P. halepensis and the consequences of pine canopy development on the performance of S. tenacissima. We hypothesised that an interaction between both species exists in alfa steppes, so that S. tenacissima facilitates the establishment of R halepensis, which then negatively affects S. tenacissima. Our results show that in steep slopes the presence of P. hatepensis individuals was heavily associated with the terracettes formed by the accumulation of sediments upslope of the tussocks of S. tenacissima. Even though the surface occupied by the terracettes was a third of the total, 72.9% of R halepensis individuals occurred in this environment. Significant differences were also found in seedling emergence (P = 0.001) and establishment (P = 0.002) of R halepensis. Both were higher in terracettes than on the bare ground among the tussocks, although no seedlings survived after the drought season. P. halepensis interferes negatively on the performance of S. tenacissima in some but not all of its life stages. The number of spikes per tussock was higher in steppes without pines and was also dependent on the tussock size (P < 0.0001). Significant differences between both environments were also found in emergence (P = 0.037) and survival (P < 0.001) of seedlings of S. tenacissima. Seedlings that emerged in the sun also developed a higher number of leaves and root biomass (P = 0.041 and 0.015, respectively). Pollination rates did not differ between the sites with pine cover and those without pine cover. These relationships may have important consequences for the dynamics and management of these plant communities. (C) 2003 Elsevier B.V. All rights reserved.