Regeneration of a marginal Quercus suber forest in the eastern Iberian Peninsula

Abstract

Question: Small and marginal forest populations are a focus of attention because of their high biodiversity value as well as the risk of population decline and loss. In this context, we ask to what extent a small, marginal Quercus suber (Cork oak) population located in the eastern Iberian Peninsula (Valencia, Spain) has the capacity for self‐regeneration and what are the factors that determine its recruitment variability. Location: Quercus suber forest in Pinet (Valencia, Spain). Methods: We performed a spatially explicit sampling both of the recruitment and of the potential parameters that could account for the recruitment variability. Using regression techniques we model the recruitment occurrence and abundance, and then we test to what extent the model obtained is still constrained by the spatial dependence. Results: Quercus suber recruitment density ranges from 0 to 18.66 individuals/25m2 (mean = 1.46, SD = 2.8), with a very skewed distribution. Recruitment is similar under Q. suber forests and under Pinus forests, but it is almost absent under shrublands. Thus the parameters that explain most of the recruitment variability in local vegetation types are: the presence and cover of shrubs (negative relationship with recruitment), the basal area of Q. suber and Pinus and the amount of bare soil (all positively related to recruitment). These parameters are strongly related to the ecological processes driving recruitment (i.e. dispersal and predation) and they remove most of the spatial dependence of recruitment. Most recruiters, however, are small, forming a seedling bank rather than growing to successfully colonize new habitats. Conclusion: The results suggest that although recruitment densities are not very high, they do not limit potential regeneration in the Pinet Q. suber forest. However, successful regeneration is not observed. If we aim to increase the Pinet Q. suber population size, land management measures need to provide appropriate conditions for both seedling establishment in shrublands (e.g. shrub clearing) and seedling growth in woodlands (e.g. Pinus logging).

Acorn dispersal estimated by radio-tracking

“…This habitat preference also concurs with the seedling regeneration densities in Mediterranean mosaic landscapes (Pausas et al 2006b;Pons and Pausas 2006). Moreover, in low-density oak populations, regeneration by rodent dispersal is usually found at a very short distance from the acorn source (Xiao et al 2005;Xiao and Zhang 2006;Pons and Pausas 2007a).…”

“…Thus, we suggest that the Quercus regeneration pattern is primarily driven by jay dispersal, at least in our mosaic landscape. However, in light of the Wnding that our jays dispersed some acorns to shrublands but that we failed to Wnd any signiWcant regeneration in this habitat (Pausas et al 2006b;Pons and Pausas 2006), some other processes could be aVecting seedling densities in this habitat (e.g. high acorn post-dispersal predation by rodents in shrublands; Pons and Pausas 2007a).…”

The coexistence of acorns with different maturation patterns explains acorn production variability in cork oak

“…Potential and actual evapotranspiration (PET and AET, respectively) was computed following Thornthwaite and Mather (1957) from monthly climatic data. For AET, we used a bucket model, with the soil holding capacity inferred from available soil analyses (i.e., textures) in the study area (for Pinet, Pausas et al 2006;for Espadà, Serrasolses et al 2009; for Calderona, Rubio et al 1995). The water deWcit (WD) was then computed as PET ¡ AET, and the drought index (DI) as DI = 1 ¡ AET/PET.…”

“…Soils in Calderona and Espadà are mainly acidic on sandstones (Bundsanstein, which is a Triassic bedrock) with soil pH ranging from 4.9 to 6.4 (mean 5.7). At Pinet, Q. suber occurs on non-carbonated soils on dolomites, with soil pH ranging from 5.5 to 7.5 (mean: 6.5; Cretaceous bedrock; see Pausas et al 2006). The three sites are currently protected either as Natural Parks (Espadà and Calderona) or as a regional plant micro-reserve (Pinet).…”

Modelling jay (Garrulus glandarius) abundance and distribution for oak regeneration assessment in Mediterranean landscapes

“…For instance, the absence of the dispersal vector will limit regeneration even when other conditions are optimal, and thus variables defining the potential regeneration niche may not be sufficiently explicative of the absence of regeneration. Accordingly, the strong heterogeneity observed in natural oak regeneration Pausas et al, 2006) may be due to not only environmental conditions but also different dispersal vector abundances in the different areas. This means that for predicting oak distribution and spread, we need to be able to predict the presence and abundance of the main oak dispersal vector.…”