Resource Communication

 

A state-and-transition model of Iberian dehesas based on spatial patterns

 

Jesús Caballero

Universidad de Extremadura, Escuela Politécnica, Avenida de la Universidad s/n 10005, Cáceres, Spain

Ángel M. Felicísimo

Universidad de Extremadura, Escuela Politécnica, Avenida de la Universidad s/n 10005, Cáceres, Spain

Elia Quirós

Universidad de Extremadura, Escuela Politécnica, Avenida de la Universidad s/n 10005, Cáceres, Spain

 

Abstract

Aim of study: In this study a process based on a Geographic Information System (GIS) is proposed as a tool for analyzing the spatial structure of the Mediterranean dehesa ecosystem.

Area of study: Western Peninsular Spain.

Material and methods: The method allows a semi-quantitative “state and transitions” (S&T) net that provides original information derived from spatial patterns that cannot be obtained by other means.

Main results: This GIS analysis also supplies a spatial basis which complements the conceptual framework of vegetation series successional proposals. Moreover, the proposed method can be a useful tool for understanding landscape changes.

Research highlights: Our work proposes not only a descriptive model but also a methodology based on spatial metrics and topological relationships.

Keywords: Geographic information systems; successional net; landscape management.

Abbreviations: S&T: State and Transition; RFP: Regional Forest Plan.

Citation: Caballero, J., Felicísimo, A.M., Quirós, E. (2015). A state-and-transition model of Iberian dehesas based on spatial patterns. Forest Systems, Volume 24, Issue 2, eRC05, 6 pages. http://dx.doi.org/10.5424/fs/2015242-06408.

Received: 14 Jan 14. Accepted: 08 Apr 2015

Copyright © 2015 INIA. This is an open access article distributed under the Creative Commons Attribution License (CC by 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Funding: The autor(s) received no specific funding for this work.

Competing interests: The authors have declared that no competing interests exist.

Correspondence should be addressed to Elia Quirós: equiros@unex.es


 

CONTENTS

Abstract

Introduction

Materials and methods

Results and discussion

Conclusions

References

IntroductionTop

Dehesas cover more than 3 million hectares of the South-West of the Iberian Peninsula (Díaz et al., 1997). Although there are different definitions for the term “dehesa”, it generally refers to managed wooded pastureland (Joffre et al., 1988, Pinto-Correia, 1993, Campos et al. 2013). They are of considerable importance from a socio-economic, environmental and cultural point of view (Campos Palacín, 1993, Díaz et al., 1997, Rensburg, 2001). In terms of sustainability, it is especially remarkable the compatibility between economic profitability and the conservational interest of the high biodiversity values (Pinto-Correia, 2000).

The dehesa landscape is characterized by a two-layered structure consisting of seasonal grassland and scattered trees dominated mainly by Quercus ilex subsp. ballota (Desf.) Samp. and in a lesser extent, Quercus suber L. and Quercus pyrenaica Willd.

While over the past they been quite stable, landscape changes have taken place in the last decades, mainly due to the mechanization of agriculture and the increase of irrigated land and reforestation with fast-growing tree species (Elena Rosselló et al., 1987, Bermejo, 1994). The remaining dehesas are reported to suffer from land degradation due to several causes: lack of trees’ regeneration (Montero et al., 1998); shrub encroachment in abandoned areas (Marañón, 1988); decrease of tree biomass due to excessive pruning for firewood production (Regato-Pajares et al., 2004); and, soil degradation because of excessive stocking rates (Schnabel, 1997). It is important to note that abandonment leads to a regeneration in the direction of what is thought to be the original vegetation cover; however, this is considered as a degradation because the dehesa is considered the suitable stage.

Most of the land use changes in the dehesa landscape take place without specific management plans, and their repercussions for the spatial patterns and the landscape are not well known. The influence of historical factors has been dealt with by (Pinto-Correia & Mascarenhas, 1999).

The correct management of the landscape involves understanding the successional patterns over medium to longer term. Spatial patterns can be a useful tool for understanding dynamic or temporal changes.

The aim of the present study is to propose a descriptive model and a methodology based on spatial metrics and topological relationships to describe a “state and transitions” net as a model of dehesa dynamics. The S&T models are nets describing changes between states: recognizable, resistant and resilient complex of soil base and vegetation structure. The above definition and a review are in (Stringham et al., 2003) and (Briske et al., 2005).

Materials and methodsTop

The study area

The study was performed in Extremadura, a region of the Iberian Peninsula with a Mediterranean climate. The dominant tree species are mainly Quercus ilex subsp. ballota, while Quercus suber and Quercus pyrenaica have also a significant presence (Figure 1).

Figure 1. Current distribution of Quercus species in the study area

Data sets

The analysis was carried out using two sets of digital maps reproducing the spatial distribution of the current vegetation. The first set was obtained from the Spanish Forestry Map (SFM). It was used in order to identify the main vegetation classes and their current spatial distribution over the study area.

The second set of maps was the suitability maps (Figure 2). The use of those maps was one of the essential points of this study. A suitability map is a raster model in which each pixel value represents the suitability for a given use (e.g. forest presence). The basis of the suitability models is to establish relationships between the environmental variables and the spatial distribution of vegetation. Commonly, each vegetation type will respond to a different model as a consequence of their different environmental requirements. The suitability is expressed on a scale of 0–1 (incompatible–ideal). A complete explanation of our suitability models can be found in (Muñoz & Felicísimo, 2004, Mateo et al., 2011) Among all techniques for generating suitability maps, multiple adaptive regression splines method (Friedman, 1991) was the selected method.

Figure 2. Example of Quercus pyrenaica (left) and Quercus suber (right) suitability models.

Spatial analysis

A Geographic Information System (GIS) analysed both the metric and topologic relationships between vegetation patches of each Quercus species.

The most suitable values, obtained from the suitability models, defined the potential area for each Quercus species as a set of patches (polygons). These polygons were intersected with the SFM in order to determine the vegetation classes actually occupying the potential areas. The surfaces were classified according to vegetation categories. The selection of the patches and their area were the basis for the design of the nets for the current vegetation states and their spatial (not temporal) transitions.

Three state and transition (S&T) nets were constructed, one for each species of Quercus, which included all the vegetation classes that have spatial relations with the tree formations. These nets were the result of the actual knowledge of the successional stages, but adapted to the observed spatial relations. Both states and transitions were evaluated by means of the statistics derived from the spatial analysis.

Results and discussionTop

State and transition nets

The S&T nets do not constitute a copy of the common succession models since they include several states that can be outside the natural succession; according to (Pivello & Coutinho, 1996), the model reflects that the succession in the study area is a multiple pathway process, which contrasts with the idea of a linear succession.

Figure 3 shows the S&T nets corresponding to the three considered species. Two types of states were considered: stationary (rectangles with round corners) and non-stationary (rectangles). The first may be possible due to natural (e.g. bedrock) or artificial (e.g. dams, irrigated lands) causes, making the change to another state difficult.

Figure 3. State and transition net for the three Quercus species.

Arrows represent the spatial structure derived from GIS analysis of the potential/neighbouring areas. The arrow thicknesses show the importance of transitions between states derived from the actual surface of the patches and they assign a semi-quantitative significance to net elements. Finally, dotted lines represent states without transitions, namely, rocky terrains (a limiting stage) close to Quercus pyrenaica.

Descriptions of stages

The following Successional stages were obtained.

S0 Forests: with a moderate degree of human intervention, equilibrated strata and a dominance of trees and shrubs.

S1 Dehesas: resulting from forest clearance and shrub elimination, with scattered trees over a continuous grass layer.

S2 Jarales: shrub formation with dominance of Cistus ladanifer L.

S3 Escobonales: shrub formation with dominance of Cytisus striatus (Hill) Rothm. and C.scoparius (L.) Link.

S4 Retamares: shrub formation with dominance of Retama sphaerocarpa (L.) Boiss.

S5 Brezales: heath with dominance of Erica australis L., E. arborea L., E. umbellata Loefl. ex L. and Calluna vulgaris (L.) Hull.

S6 Grasslands: these herbaceous formations can only be considered a natural stage in some mountainous areas, where there are Nardus stricta L. domains; the floristic composition is strongly dependent on the climatic and edaphic conditions.

On the other side, the obtained Substitution stages were the following:

S7 Crops: tillage on irrigated lands. This stage can be interpreted as a stationary state in the S&T net.

S8 Chestnut: in Extremadura only, appearing in the most humid parts of the mountainous areas.

S9 Pinus pinaster Aiton and Pinus pinea L.: the most abundant species of the Pinaceae family; they are exotic in the study area.

S10 Eucalyptus: mainly restricted in Extremadura to the humid areas, near to the rivers; E. camaldulensis Dehnh. and E. globulus Labill. are the most abundant species.

The Limiting stage (L0) can be interpreted as stationary in the S&T net. In the highest areas, the ridges of bare rock constitute the physical limit to the growth of vegetation.

Description of transitions

Progressive transitions in the model were the following:

T1 Dehesa to forest: This transition could be considered as “natural” due to the successional instability of the dehesas. Inverse transitions, from forest to dehesa, probably do not exist nowadays.

T2 Dehesa to dehesa loops: The control of shrub encroachment has been traditionally carried out by direct interventions. Abandonment of these cultural practices decreases the global weight of T2 in favor of T1.

T3 Grassland tojaral: This transition is present on the three Quercus nets. Cistus is commonly interpreted as a pioneer shrub species which quickly colonizes areas of degraded soils after some natural or human induced perturbations.

T4 Jaral to other shrub formation. The Cistus roots break the ground and their litter contributes by increasing the content of organic matter and improving the soil texture.

T5 Dense shrubbery to early forest. This implies the development of a tree layer. This process is possible only if trees exist close to the space of the shrub formation, and it is unlikely to occur in other places.

On the other hand, the obtained regressive transitions were:

T6 Dehesas or forests to other timber species or croplands: Croplands, mainly on the irrigated lands, are quasi-stationary stages since their reversion to any other natural state is actually impossible. This is not the case for timberlands, because the reduction of the human interventions can contribute to colonization by Mediterranean shrub species.

T7 Pinus to shrub formation. The mosaic of patches of Pinus suggests a facilitated transition to escobonales (Cytisus sp.) which could be originated by timber production activities. It seems probable that succession of these areas leads to a dense shrub stratum.

ConclusionsTop

We propose S&T models as a tool for analyzing the spatial structure of the Mediterranean dehesa ecosystem. The S&T models provide original information derived from spatial patterns and complement the conceptual framework of successional proposals for vegetation series. Moreover the S&T models are relatively simple compared to predictive or biologically-based models (Rogers & Johnson, 1998, Lambin et al., 2000).

The interest of the proposed models lies in their ability to provide objective data about the spatial and topological relationships between vegetation patches and establish a conceptual framework for the dynamic of vegetation categories.

Obviously, the S&T models should be discussed carefully in each case since the model assumptions can be appropriated only under some circumstances. The dehesa ecosystem has evolved influenced by natural and anthropogenic factors under common goals and constraints: the optimization of the land uses under strong climate conditions and avoiding expensive inputs as irrigation or artificial fertilization. This forces to optimally exploit the natural resources of the land and conditioned the potential stages and transitions. We believe that in other circumstances (friendly climate and intensive artificial inputs) the S&T models can be less significant.

With these limitations, the advantages of the S&T models are their objectivity and repeatability. Nevertheless, any statistical or GIS model is heavily dependent on data and it would have been very advantageous to have historical data to analyze the sequence with some intermediate states.

It is necessary to emphasize that the actual patterns are only a trace of the history but we believe that this work shows the potential of the S&T nets as a simple but objective technical to help address some management actions.


ReferencesTop

Bermejo I, 1994. Conservación de sistemas adehesados en Extremadura. Agricultura: Revista agropecuaria 738: 40-43.
Briske DD, Fuhlendorf SD,Smeins FE, 2005. State-and-Transition Models, Thresholds, and Rangeland Health: A Synthesis of Ecological Concepts and Perspectives. Rangeland Ecol Manag 58(1): 1-10.
Campos Palacín P, 1993. Valores comerciales y ambientales de las dehesas españolas. Agricultura y sociedad 66: 9-42.
Campos P, Huntsinger L, Oviedo JL, Starrs PF, Diaz M, Standiford R, Montero G, 2013. Mediterranean oak woodland working landscapes: Dehesas of Spain and Ranchlands of California, Springer Science Business Media. http://dx.doi.org/10.1007/978-94-007-6707-2
Díaz M, Campos P, Pulido F, 1997. The Spanish dehesas: a diversity in land-use and wildlife. Farming and birds in Europe: the common Agricultural Policy and its implications for bird conservation. Pain and Pienkowski. London, Academic Press: 178-209.
Elena Rosselló M, Lopez Marquez J, Casas Martín M, Sánchez del Corral Jimenez A, 1987. El carbón de encina y la dehesa. Instituto Nacional de Investigaciones Agrarias, Madrid (Spain).
Friedman JH, 1991. Multivariate adaptive regression splines. Ann Stat 19(1): 1-67. http://dx.doi.org/10.1214/aos/1176347963
Joffre R, Vacher J, de los Llanos C, Long G, 1988. The dehesa: an agrosilvopastoral system of the Mediterranean region with special reference to the Sierra Morena area of Spain. Agroforest Syst 6 (1-3): 71-96. http://dx.doi.org/10.1007/BF02344747
Lambin EF, Rounsevell M, Geist H, 2000. Are agricultural land-use models able to predict changes in land-use intensity? Agr Ecosyst Environ 82(1): 321-331. http://dx.doi.org/10.1016/S0167-8809(00)00235-8
Marañón T, 1988. Agro-sylvo-pastoral systems in the Iberian Peninsula: Dehesas and Montados. Rangelands: 255-258.
Mateo RG, Felicísimo AM, Muñoz J, 2011. Modelos de distribución de especies: una revisión sintética. Rev Chil Hist Nat 84: 217-240. http://dx.doi.org/10.4067/S0716-078X2011000200008
Montero G, San Miguel A,Cañellas I, 1998. Sistemas de silvicultura mediterránea. La dehesa. Agricultura sostenible. R. M. Jiménez Díaz and J. Lamo de Espinosa. Madrid, Spain, Mundi-Prensa: 519-554.
Muñoz J, Felicísimo AM, 2004. A comparison between some statistical methods commonly used in predictive modeling. J Veg Sci 15: 285-292. http://dx.doi.org/10.1111/j.1654-1103.2004.tb02263.x
Pinto-Correia T, 1993. Threatened landscape in Alentejo, Portugal: the ‘montado’ and other ‘agro-silvo-pastoral’ systems. Landscape Urban Plan 24(1–4): 43-48. http://dx.doi.org/10.1016/0169-2046(93)90081-N
Pinto-Correia T, 2000. Future development in Portuguese rural areas: how to manage agricultural support for landscape conservation? Landscape Urban Plan 50(1): 95-106. http://dx.doi.org/10.1016/S0169-2046(00)00082-7
Pinto-Correia T, Mascarenhas J, 1999. Contribution to the extensification/intensification debate: new trends in the Portuguese montado. Landscape Urban Plan 46(1): 125-131. http://dx.doi.org/10.1016/S0169-2046(99)00036-5
Pivello VR, Coutinho LM, 1996. A qualitative successional model to assist in the management of Brazilian cerrados. Forest Ecol Manag 87(1): 127-138. http://dx.doi.org/10.1016/S0378-1127(96)03829-7
Regato-Pajares P, Jiménez-Caballero S, Castejon M, Elena-Rosselló R, 2004. Recent landscape evolution in Dehesa woodlands of western Spain. Recent dynamics of the Mediterranean Vegetation and Landscape. S. Mazzoleni, G. D. Pasquale, M. Mulligan, P. D. Martino and F. Rego. Chichester, UK, Wiley: 57-72.
Van Rensburg TM, 2001. The Spanish dehesas: a financial appraisal of holm oak (Quercus ilex) regeneration. Forest Systems 10(3): 125-136.
Rogers R, Johnson PS, 1998. Approaches to modeling natural regeneration in oak-dominated forests. Forest Ecol Manag 106(1): 45-54. http://dx.doi.org/10.1016/S0378-1127(97)00238-7
Schnabel S, 1997. Soil erosion and runoff production in a small watershed under silvo-pastoral landuse (dehesas) in Extremadura, Spain. Logroño, Spain, Geoforma.
Stringham TK, Krueger WC, Shaver PL, 2003. State and transition modeling: An ecological process approach. J Range Manage 56: 106 -113. http://dx.doi.org/10.2307/4003893



Webpage: www.inia.es/Forestsystems