Oak regeneration in heterogeneous landscapes: The case of fragmented Quercus suber forests in the eastern Iberian Peninsula

https://doi.org/10.1016/j.foreco.2006.05.049Get rights and content

Abstract

Quercus suber recruitment is quantified at local (in different vegetation types) and landscape level in three marginal populations (sites) ranging in size from 70 to 7000 ha and located in the eastern Iberian Peninsula (Spain). We hypothesised that: (1) recruitment of marginal Q. suber populations in eastern Iberia should be higher than in the core area of distribution, western Iberia, (2) within our study area, there should be a trend in the regeneration from north (the largest forest patch) to south (the smallest patch), and (3) within a site, recruitment should not be randomly distributed, but rather some vegetation types should show higher recruitment than other vegetation types. To test these hypotheses, a total of 61 plots measuring 12.5 m × 12.5 m were established on seven vegetation types in five previously selected 3 km × 3 km quadrats. On each plot all Quercus plants shorter than 3 m were labelled and tracked between 2003 and 2005. Results support the first and third hypotheses but not the second one. Q. suber recruitment densities ranged from 0 to 7200 plants per hectare. The complete regeneration process (from recruitment to growth) occurred on old fields; well-established seedling banks were abundant under forest canopies; recruitment in shrublands was practically null. To better understand Q. suber recruitment, both landscape configuration and interactions with other Quercus species should be considered.

Introduction

Natural oak regeneration is the product of complex processes such as acorn production, predation, animal dispersion, germination and establishment in the habitat where the acorns were dispersed, and finally, growth to mature trees (Schupp, 1990). The combination of all these processes may determine the success or failure of oak regeneration and the final fitness of this species to a given set of conditions. Thus, evaluating natural regeneration is an integrative way to evaluate fitness and provides a first indication of the future of the populations under study.

Quercus suber (cork oak) is an evergreen tree species from the western part of the Mediterranean Basin (Tutin et al., 1964). In the Iberian Peninsula, most Q. suber occur in a highly managed savannah-like agrosystem (called dehesa or montado) on acidic soils (the core area). Nevertheless, several patches of cork oak (peripheral cork oak patches) also occur in the eastern part of the peninsula, where calcareous soils are dominant. There are some large Q. suber forests in the northern part of the eastern Iberian Peninsula (Catalonia, ca. 70,000 ha, Montero et al., 1994); however, patch size decreases towards the south by about three orders of magnitude. Thus, in the Valencia region (eastern Iberia), the main Q. suber forest patches range from ca. 70 to 7000 ha (Pausas et al., 2004) and are considered marginal and fragmented, and their future has been questioned. Even so, there is still some cork industry in the region. In fact, cork oak forests are protected by the European Union (Habitat directive 92/43/EEC), and the cork oak is considered an important species for the socio-economic development of some European regions due to its potential use in sustainable forestry (Q. suber products, cork and acorns, do not require the elimination of trees but rather their protection). Furthermore, Q. suber is a very good candidate for reforestation plans in the Mediterranean area due its efficient and rapid post-fire regeneration capacity (Pausas, 1997). All these aspects contribute to the importance of amassing knowledge on this species and its natural regeneration capacity in order to focus further restoration plans (Mosandl and Kleinert, 1998, Pausas et al., 2004). To date, however, very little research has been performed on the regeneration of Q. suber (Montero et al., 1994), and it has been carried out only in the core area of the species distribution. There is increasing evidence of the conservation and biodiversity value of peripheral populations (Lesica and Allendorf, 1995, Hampe and Petit, 2005), and specifically of Q. suber, due to the distinct genetic pools between central and peripheral populations (Jiménez et al., 1999).

Most current landscapes are highly heterogeneous at several spatial scales, and studies on oak dispersal and regeneration should account for this spatial variability in order to understand the future of this species. Thus our initial hypotheses are framed within different spatial scales: between eastern and western Iberian Peninsula, between different sites of the eastern Peninsula, and between different patch types within the landscape.

Regeneration problems have been described for a variety of Quercus species (e.g., Shaw, 1968, Mellanby, 1968, Loftis and McGee, 1993, Lorimer et al., 1994), but little is known about regeneration of Mediterranean species. Part of the research on Mediterranean Quercus has been performed in the dehesa agrosystems of SW Spain (Plieninget et al., 2004, Pulido and Díaz, 2005). In such dehesas, oak regeneration seems to be much lower than in forest ecosystems (Pulido and Díaz, 2005), probably because intense human impacts (Plieninget et al., 2004) have resulted in an unsuccessful acorn dispersal and a heavy grazing pressure (Pulido and Díaz, 2005). In eastern Iberia, these pressures (intense human impact and grazing) have decreased in recent last decades due to rural exodus and abandonment processes occurring on European Mediterranean coasts (Pausas, 2004). Thus, our first hypothesis is that Q. suber regeneration in eastern populations will be higher than Quercus regeneration in dehesas.

Our second hypothesis is that regeneration will decrease from northern (larger) to southern (smaller) populations (site–scale analysis) and that this may be due to different processes. It is well known that small populations may face both inbreeding effects and genetic erosion, which result in a reduction in their viable seed production and seed quality and make them more prone to extinction (Lawton, 1993, Vucetich and Waite, 2003). This would be reflected in the regeneration process. In this context, it has been demonstrated that marginal Q. suber populations have lower genetic diversity than central populations (Jiménez et al., 1999). Thus, we could expect a decreasing regeneration pattern that parallels the size of the different populations, especially with respect to the southernmost population (Pinet), which is considered small for a tree population (ca. 70 ha of low-density trees). Furthermore, the fact that forests are the main habitat for the oak-dispersal agent (the European jay Garrulus glandarius) and there is a decreasing abundance of forest stands from north to south in our study area (Pausas et al., 2004) could also contribute to reduced Q. suber regeneration. Finally, small populations have a higher edge effect, which may imply higher predation (Santos and Telleria, 1997) and disturbance sensitivity. All these processes would suggest that regeneration would decrease from northern to southern Q. suber populations in eastern Iberia.

Our third hypothesis is that, at landscape scale, regeneration will not be randomly distributed because environmental conditions and biotic factors (dispersal, predation, competition) affecting seed and seedling success vary in different landscape units (Grime and Hillier, 2000, Gómez et al., 2004). Thus, given the current heterogeneity of our landscapes we expect differential recruitment and regeneration processes for different spatial units.

To test these hypotheses we quantified the amount and variability of Q. suber recruitment and that of all other Quercus species appearing within the study area in different landscape units around the three main Q. suber patches in the Valencia region. We considered the following non-sequential stages of regeneration: recruitment, development of seedlings, formation of a seedling (sapling) bank and development to mature trees. The final objective was to understand to what extent these fragmented marginal cork oak populations are stable and would persist in the study area. Alternatively, other oak species may preferentially occupy the potential habitat of Q. suber and thus displace it.

Section snippets

Study area

The study area is located inside the Valencia region, which borders the Mediterranean coast in eastern Spain (Fig. 1). The climate is typically Mediterranean. The vegetation is a product of a long history of fire and land use, and many slopes were terraced and cultivated in the past, and then abandoned. In this region there are three main Q. suber patches (hereafter sites), from north to south: Espadà (Castelló, lat = 39°52′, ca. 7,000 ha), Calderona (València, lat = 39°44′, ca. 700 ha) and Pinet

Landscape analysis

Landscape pattern differs between different quadrats and sites (Table 2). The northern quadrats (Espadà) are more forested than the Calderona and Pinet quadrats. However, the type of forest differs between the two Espadà quadrats: Quadrat 1 includes a mixed forest with several oak and pine species, while Quadrat 2 is the result of several recent fires and the only dominant tree species is Q. suber. Quadrat 3 (Calderona) is the most patched quadrat, with both forests and shrublands. Quadrat 4

Cork oak recruitment pattern

To our knowledge this is the first paper studying Q. suber recruitment at landscape level, and thus, most of the comparisons we make hereafter will be with other Quercus species, especially Q. ilex.

Overall Q. suber recruitment densities in marginal cork oak populations of the eastern Iberian Peninsula are highly variable depending on the vegetation type considered, and they fall within the broad range of density values available for Quercus ilex in nearby areas (see Table 6). In this region the

Conclusions

Overall Quercus recruitment in the eastern Iberian peninsula is higher than in the western Iberian dehesas. Q. suber recruitment is relatively high in old fields and forest of the study area, at least in the range of distances from adult cork oak trees within the jay dispersal distances range; but recruitment is absent in shrublands. However, successful regeneration only occurs in old fields, while in forests (in both pine and oak forests), seedling development is suppressed by adult trees, and

Acknowledgements

We thank Cristina Betseller, Eloi Ribeiro, Nestor Pons, Marcos Mata, Ana B. Ribes, and Alicia Roman for collaborating in the field sampling. This work was funded by the CREOAK European project (QLRT-2001-01594) and the VARQUS Spanish project (CGL2004-04325/BOS). CEAM is supported by Generalitat Valenciana and Bancaixa.

References (46)

  • J.M. Espelta et al.

    Pattern of seedling recruitment in West-Mediterranean Quercus ilex forests influenced by canopy development

    J. Veg. Sci.

    (1995)
  • I. Frost et al.

    Spatial pattern and size distribution of the animal-dispersed tree Quercus robur in two spruce-dominated forests

    Ecoscience

    (2000)
  • L.O. George et al.

    The fern understory as an ecological filter: Growth and survival of canopy-tree seedlings

    Ecology

    (1999)
  • J.M. Gómez

    Spatial patterns in long-distance dispersal of Quercus ilex acorns by jays in a heterogeneous landscape

    Ecography

    (2003)
  • J.M. Gómez et al.

    Differences between structural and functional environmental heterogeneity caused by seed dispersal

    Funct. Ecol.

    (2004)
  • M. Gracia et al.

    Seedling bank dynamics in managed holm oak (Quercus ilex) forests

    Ann. For. Sci.

    (2001)
  • J.P. Grime et al.

    The contribution of seedling regeneration to the structure and dynamics of plant communities, ecosystems and larger units of the landscape

  • A. Hampe et al.

    Conserving biodiversity under climate change: the rear edge matters

    Ecol. Lett.

    (2005)
  • P.E. Hulme

    Post dispersal seed predation in grassland: its magnitude and sources of variation

    J. Ecol.

    (1994)
  • P.E. Hulme

    Post-dispersal seed predation and the establishment of vertebrate dispersed plants in Mediterranean scrublands

    Oecologia

    (1997)
  • P. Jiménez et al.

    Genetic variation in central and marginal populations of Quercus suber L.

    Silvae Genet.

    (1999)
  • W.C. Johnson et al.

    Nut caching by blue jays (Cyanocitta cristata L): Implications for tree demography

    Am. Midl. Nat.

    (1997)
  • D. Kelly et al.

    Mast seeding in perennial plants: why, how, where?

    Annu. Rev. Ecol. Syst.

    (2002)
  • Cited by (83)

    View all citing articles on Scopus
    View full text