SciELO - Scientific Electronic Library Online

vol.37 número2Identificación de las proantocianidinas extraídas desde corteza de Pinus Radiata D. Don índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados


Ciencia e investigación agraria

versión On-line ISSN 0718-1620

Cienc. Inv. Agr. v.37 n.2 Santiago ago. 2010 

Cien. Inv. Agr. 37(2):5-14. 2010



Genetic parameters of growth and survival in Acacia saligna shrubs

Parámetros genéticos del crecimiento y sobrevivencia, en arbustos de Acacia saligna


Freddy Mora1, Sandra Perret2, Carlos Alberto Scapim3, and Emmanuel Arnhold4

1Facultad de Ciencias Forestales, Universidad de Concepción, Victoria 631, Barrio Universitario, Concepción, Chile.
2Instituto de Investigación Forestal, Ministerio de Agricultura de Chile, Huérfanos 554, Santiago, Chile.
3Centro de Ciencias Agrarias. Universidade Estadual de Maringá. Avenida Colombo, 5790, Bloco 5 - PGM, Maringá-PR, Brasil.
Escola de Veterinaria, Universidade Federal de Goiás, Campus II, Samambaia, Caixa Postal 131, CEP: 74001- 970, Goiânia, Goiás, Brasil.

Dirección para correspondencia


Acacia species have valuable agronomic characteristics, such as the tolerance to unfavorable environmental conditions. In this study, genetic parameters and selection response in Acacia saligna shrubs were investigated in two arid environments of northern Chile: Cuz-Cuz and El Tangue. The quantitative traits were assessed in 5 year-old shrubs. A Bayesian approach, implemented using Gibbs sampling algorithm, was used in the data analysis. The tree survival was 87.3% and 81.8% in Cuz-Cuz and El Tangue, respectively, confirming the potential of A. saligna in northern Chile. The estimates of posterior mode of the heritability ranged from h2=0.13 (survival) to h2=0.25 (shrub crown diameter). Estimated genetic gain (individual tree selection) ranged from 6 to 14% (selection intensity of 20%). Bayesian credible intervals (p=95%) for genotype-environment correlations included zero, indicating a significant interaction for survival, crown diameter and height. Shrub selection in multipurpose planting will depend on optimizing economical traits by selecting genotypes that perform well on a particular environment.

Key words: Drylands, fodder production, Acacia saligna, multipurpose trees, Bayesian analysis.


Las especies de acacia poseen características valiosas, como la tolerancia a condiciones ambientales desfavorables. En el presente estudio, se investigaron parámetros genéticos y la respuesta a selección en arbustos de Acacia saligna, en dos ambientes áridos del norte de Chile: Cuz-Cuz y El Tangue. Las características cuantitativas se evaluaron en arbustos de 5 años de edad. Para el análisis de datos se utilizó un enfoque Bayesiano, implementado usando el algoritmo de Gibbs. La sobrevivencia fue de 87,3 y 81,8% en Cuz-Cuz y El Tangue, respectivamente, confirmando el potencial de A. saligna en el norte de Chile. Las estimaciones de la moda a posteriori de la heredabilidad varió de h2=0,13 (sobrevivencia) a h2=0,25 (diámetro de copa). La ganancia genética (selección de árboles individuales) varió de 6 a 14% (intensidad de selección de 20%). Intervalo de credibilidad Bayesiano (p=95%) de las correlaciones genotipo-ambiente incluyeron el valor cero, indicando una significativa interacción para las características de sobrevivencia, diámetro de copa y altura. La selección de arbustos en plantaciones multipropósito, dependerá de la optimización de las características de interés a través de la selección de genotipos que responden bien en un sitio en particular.

Palabras clave: Zonas áridas, producción de forraje, Acacia saligna, árboles multipropósito, análisis Bayesiano.


The species Acacia saligna (ex Acacia cyanophylla) is considerad a tree resource of known aptitude for the establishment of agroforestry systems in arid environments (Hnatiuk and Maslin, 1988; McDonald and Mashn, 2000).

Some physiological traits of A. saligna have favored the development of some mechanisms for drought tolerance (Nativ et al., 1999; Tiedeman and Johnson1992) allowing the settlement of multipurposepiantations under adverse conditions of climate and soil (Degen et al. 1995; Nativ, et al. 1999; Midgley and Turnbull 2003; Mora and Meneses 2004; George et al., 2006).

This species has been planted for forage and wood production, as ornamental plant,and for land rehabilitation and soil conservation (Degen et al., 1995; McDonald and Maslin, 2000). In arid ecosystems, the settlement of mixed systems of cultivation using trees or shrubs of A. saligna is useful for soil protection, allowing nitrogen fixation, providing wood and forage during the dry season (Tiedeman and Johnson, 1992).

In arid zones, primary productivity varies widely from an almost nuil productivity in extremely arid ecosystems, to values approaching dry tropical deciduous forests (Burquez et al., 1999). Arid environments have a wide biodiversity of flora and fauna, and a heterogeneous landscape, but they have similar primary productivity due to the wide variation of rain and high rates of evapotranspiration, with abundant periods of drought (Wenninger and Inouye, 2008). These adverse conditions also occur in several countries of South America with continuous periods of drought, increasing soil salinity and desertification (Holmgren et al., 2006).

In Chile, the administrative region of Coquimbo presents a proved potential for the development of large scale plantations of the species. It corresponds to a Mediterranean type zone with arid tendency, and it is located in the southern part of the Atacama Desert. According to Squeo et al. (2006) this Region presents critical productivity due to the continuous drought periods, which are affecting the social and economic activities of farmers; although previous studies have emphasized the fact that it is an appropriate Region to establish silvopastoral systems, improving the property productivity of agricultural systems (Ormazábal, 1991). Based on this principie, the Chilean Ministries of Agriculture and Economy have financed important projects with this species (Mora and Meneses, 2004; Meneses and Flores, 1999), as well as projects based on other forest alternatives associated to this cultivation (Mora and Perret, 2007; Mora et al., 2007). The Chilean governmental agencies have projected a potential surface of more than a million of hectares for plantations with this species; most of them susceptible to be covered with the Law Decree 701 for forest foster (Mora et al., 2004).

The establishment of provenance and progeny test trials is the first step for improvement programs of forest trees, which allow the identification of the best resources for soil rehabilitation and agroforestry production (Velarde et al., 2003). The genetic variability in genotypes of A. saligna is a key factor to improve productivity and survival in drought periods (George et al., 2006; Mora and Meneses, 2004).

The widest genetic diversity of A. saligna occurs in the Mediterranean Region in the South West of Australia, with a regime of precipitations varying between 0 (a 250) mmxyear1 in the central part, to 1000 mmxyear1 in the South Region (Howard et al., 2002). Provenances of forest trees with a wide geographic distribution generally show a significant variation in their anatomic, morphological and physiological traits. This genetic variability among populations may be researched ex-situ using agroforestry tests, which allow the identification of superior trees for múltiple purposes (Beniwal et al., 1995). The forest plantation programs are usually intended for high biomass productivity, combining acceptable rates of growth and survival. Subsequently, the survival of trees in juvenile states of development has been a trait of interest in different programs of selection and improvement (Coutinho et al., 2004; Mora et al., 2009; Rodovalho et al., 2008).

The analytic objective of the present study was to examine genetic parameters of growth and survival for Acacia saligna shrubs, using the Bayesian inference, based on two genetic tests established for agroforestry rehabilitation and production, under the arid conditions in the South of the Atacama Desert, North of Chile.

Materials and methods

Description of the test and sites

This study was carried out in two sites located in Southern Atacama Desert, in the region of Coquimbo (Figure 1). The site El Tangue is located in the Province of Elqui ß0°45' SL; 71°47' WL) with an average temperature of 14 °C; on the other hand, the site Cuz-Cuz is located in the Province of Choapa ß1°63' SL; 71°22' WL), with an average temperature of 15 °C. The sites (Figure 1) were selected based on the climatic conditions and on the previous experience of farmers cultivating A. saligna. The average re-gime of precipitation is frequently in the range of 100 and 200 mm year -1 in El Tangue and greater than 200 mm year -1 in Cuz-Cuz. The records of precipitation during the experimental period (1999-2004) were: average of 121.6 and 217.6 mm year-1, with minimum levels of 65 and 103 mm year-1, and maximum levels of 192 and 394 mm year-1, for El Tangue and Cuz-Cuz, respectively.

The trials were hand-planted with 14 provenances in both sites using a completely randomized block design, with trees spaced at 2 x 3 m. In Cuz-Cuz, the shrubs were grazed (overgrazed) by goats in early summer, in the last two years of experiment.

Vegetal material and quantitative traits

The provenances correspond to some of the base population maintained by CSIRO (CSIRO Australian Tree Seed Centre), which are from different localities of Australia (natural distribution of the species). The location of each population is shown in Table 1. The shrubs of A. saligna were measured at five years old considering: canopy diameter (CD), height (H), basal diameter (BD), and survival (measured as a binary response).

Statistical analysis andparameters estimation

The linear model representing the data from the experiment was the following:

This corresponds to a bi-character model, where Y1 and Y2 are the vectors of observations for a specific trait in the environments 1 and 2, respectively, considering the effect ofthe genotype-environment interaction. Xis the incidence matrix for the vector ß, which represents a vector with the Block and Site effects and associated interactions. Z is the incidence matrix ofthe vector of additive effects ofthe individual plants a. P is the incidence matrix ofthe effects vector, due to the provenance p. ε is the residual vector. The (co)variance components, heritability sensu stricto, and the prediction of the additive effects and provenances were researched by using Bayesian inference, implemented by means ofthe Gibbs algorithm.

With the Bayesian model, the variance matrices of genetic, provenance and residual effects have the following a priori distribution, respectively:

where, G0; P0 and R0 are the matrices with the a priori values of the components of (co) variance. v_ is the degree of Bayesian credibility and m_ is the dimension of the scale parameter matrix. The a priori distribution of ß was considered uniform (constant) and the vectors a and p normally distributed:

The synthesized aposteriori distribution is:

In the case of survival, the statistical analysis considered the threshold model (Van-Tassell et al., 1998), whose model implemented corresponded to:

where η is the vector of a continuous randomized variable, normally distributed, non observable (unknown), and corresponding to the variable survival (in this case bi-character).

The program MTGSAM and its Threshold version (Van-Tassell and Van-Vleck, 1996) were used to obtain the further densities of the parameters of interest (Van-Tassell et al., 1998).

The initial size of the chain was determined in 500,000 samples, with withdrawal intervals to each 40 samples and a conservative period of initial discard (burn-in) of 100,000 samples of Gibbs. The seasonability of the Gibbs chains (convergence) was evaluated through the method of Heidelberger and Welch (1983), available in the statistical program R-CODA (Cowles and Carlin, 1995).

The a posteriori heritability or degree of additive genetic control was calculated in each test and for each trait, according to the following relation:

where , correspond to the mode values of the a posteriori distributions of the additive, provenance and residual variances.

The additive genetic association (relation) for the same trait measured in two environments was calculated as:

where σaxy the mode of the posterior distribution of the additive covariance is calculated betweenthecharacteristics Xe Y. are the mode values of the a posteriori distributions of the additive variances of the traits X and Y. An r value close to 1 indicates a scarce genotype-environment interaction, while a value under r (not significantly different from zero) indicates the significant existence of genotype-environment interaction, and the ranking may be different for each site (Sykes et al., 2006).

The predicted genetic gains were estimated using the genetic values of the individual trees (breeding values) according to Mora et al. (2009). Ranking with the genetic values were made and the associations betweenthe rankings of each trait by means of the Spearman correlation coefficient, were measured.

Results and discussion

The mode values and the regions of credibility (95% of probability), of the a posteriori distributions of the components of variance and heritability are shown in Table 2. In the Bayesian context, the further distribution of the genetic +parameters may be characterizedby the median, mean and mode, which are considered as values of central tendency, and are potential point estimators. In the present study, the mode of the distributions of the parameters was considered as point measurement, according to Wright et al. (2000), who recommended this estimator to describe the further distribution. This is the region of high density, that is, the value maximizing the function ƒh2|y, or the most probable value of h2 (Blasco, 2001).

The magnitude of the values of heritability was moderate. The canopy diameter and the tree height had the highest values of heritability in El Tangue, while the canopy diameter had the highest magnitude in Cuz-Cuz. The heritability values of survival were similar in both sites and presented the lowest magnitudes.

The genetic variability confirms high possibilities to improve traits of growth and forage productivity. That information is important for the start of a small-scale breeding program for the species in southern Atacama Desert. Agroforestry programs in arid zones using A. saligna, consider the physiological mechanisms inducing the susceptibility of the plants of water déficit (Nativ et al., 1999; McDonald and Maslin, 2000). Although A. saligna is highly tolerant to dry conditions (Nativ et al., 1999), the variability found for survival may be used to improve this tolerance to dry conditions.

The genetic variability found in the present study is coherent with the studies by George et al. (2006), who made a genetic analysis by molecular markers, and they found a large varia-tioninA saligna, although two groups of plants morphologically different had the same genetic profile; this high variation between and within groups of trees suggests a wide genetic base, useful for the improvement purpose. In the present study, the genetic variability between individuals may be used for purposes of selection in both localities, particularly regarding growth.

Survival is an important trait in the agroforestry systems with water déficit (Abebe, 1994). In the present study, the shrubs showed a high survival in both sites, with averages of 87.3 and 81.8% for Cuz-Cuz and El Tangue, respectively. These results are coincident with Abebe (1994), who compared the development of different multi-purpose species appropriate for desertic zones, finding that the A. saligna shrubs had the best rates of survival, plant height, and trunk diameter under semiarid conditions in the Southern Ethiopia. In the present study, both sites are under the influence of climatic conditions affecting the competency for water resource and other limiting factors. Not only the plants of A. saligna may be affected by water shortage. According to Jara et al. (2006), in an El Niño southern oscillation event (ENSO; increased precipitation in relation to the average) the seeds from local plants that are dormant may germinate and cause a strong competency for water, nutrients, and light. 2002 was an ENSO year (Montecinos and Aceituno, 2003).

The survival of the trees in different environmental conditions may be affected by temporary or long frosts, droughts, wind, competency, and disease. The influence of these factors on mortality and precise prediction of the genetic effects through tests, established in different environments, the magnitude of the genotype-environment interactions for survival must be known in advance and studied for obtaining the maximum productivity in the site (Chambers et al., 1996). Successful improvement programs in arid zones (Otieno et al., 2001) depend on the previ-ous selection of provenances already adapted to the local conditions of climate, soil and climatic changes (Broadmeadow et al., 2005). The best provenances of A. saligna, determined according to the Bayesian prediction, are shown in Table 3. The magnitude of the provenance-environment interaction depended on the trait. The interaction was significan! according to Bayesian credibility intervals, for the canopy diameter, basal diameter and survival. In both sites, the provenances of Murchison River and Greenough River had a high prediction for the canopy diameter; a trait significantly correlated with forage prediction (Mora and Meneses, 2004), although Greenough River presented a low value predicted for survival in El Tangue.

Murchison River in El Tangue presented a high value of prediction for the canopy diameter.

However, the Bayesian credibility interval of the intraclass correlation (provenance level) indicated a significara provenance-site interaction for that trait. The genotype-environment interaction was also evident in the genetic correlations determined at level of individual trees (Table 2). The Bayesian credibility intervals also included the value Zero of the additive genetic correlation (calculated between environments) for the traits of canopy diameter, height and survival.

The values of genetic gain varied between 6% (basal diameter) and 14% (survival) (Table 4), showing that importara advances may be obtained in the process of selection. The annual harvest of A. saligna shrubs reduces the amount of forage for cattle, wood production, and may also reduce vigour, productivity, and life span of the tree (Tiedeman and Johnson, 1992). These responses may vary between genotypes, which is a significara fact for creating a program of genetic improvement under arid conditions of the Atacama Desert.

The Spearman correlation coefficients between the ranking of growth characteristics were positive and significantly different from zero (determined by the Bootstrap confidence interval, Table 4). The Spearman coefficients between the traits of growth and the rate of survival varied between -0.06 and 0.01, which were not significantly different from Zero. These non-significant correlation coefficients indicate that the selection of shrubs, based only on high levels of survival, would not have a positive and significara effect on growth (and viceversa). For example, the provenance of Ravensthorpe 2 in Cuz-Cuz was the best response of prediction (breeding value) for the canopy diameter, but not for survival.

Understanding growth and mortality of trees, and how they affect productivity of silvipastoral systems, is fundamental for the success of any pro-gram of agroforestry improvement (Chambers et al., 1996). The results of the present study are consistent with the finding of Schneider et al. (2005). who found that mortality of other species of the Acacia genus was highly correlated to the traits of growth, with values of Pearson correlation coefficients varying from 0.49 to 0.73. In the present study, survival had a weak association with growth (measured by the rankings); although high values of prediction for Murchinson River in El Tangue, indicate the possibility of joint selection mechanisms betweenboth traits. This result is rel-evantforthis improvement program with A saligna, as the genetic association between the growth rate and resistance to drought tends to be scarce, according to the studies by Coyle et al. (2006) and Schneider et al. (2005).

The genetic variability existing among these breeding populations may be used for selection and improvement purposes of traits of interest for the arid conditions of the North of Chile. It is suggested that the selection processes are not based only on growth, as it may affect the progress of other features of interest for the genetic program. In this regard, it would also be relevant to consider additional traits, such as the nutritional value of forage, improvement in the capacity of nitrogen fixation, or an increased capacity of capture of carbon dioxide, in order to increase productivity in continuous periods of drought. There is a genotype-environment interaction and it must be considered in the current program, by the selection of individuals (or genotypes) responding well in a particular a site.



Abebe, T. 1994. Growth performance of some multipurpose trees and shrubs in the semi-arid areas of Southern Ethiopia. Agroforestry Systems 26:237-248.        [ Links ]

Beniwal, R.S., O.R Toky, and RK. Sharma. 1995. Genetic variability in symbiotic nitrogen fixation between provenances of Acacia nilotica (L.) Willd. ex Del. Genetic Resources and Crop Evolution 42:7-13.        [ Links ]

Broadmeadow, M.S.J., D. Ray, and C.J.A. Samuel. 2005. Climate change and the future for broad-leaved tree species in Britain. Forestry 78:145-161.        [ Links ]

Chambers, P.G.S., N.M.G. Borralho, and B.M. Potts. 1996. Genetic analysis of survival in Eucalyptus globulus ssp. Globulus. Silvae Genética 45:107-112.        [ Links ]

Coutinho, J.L.B., V.R Santos, R.L.C. Ferreira, and J.C.B. Nascimento. 2004. Performance of Eucalyptus spp. seedlings in the "dry rainforest" region of the State of Pernambuco, Brazil. I: Results of the first year. Rev. Arvore 28(6):771-775.        [ Links ]

Coyle, D.R., M.D. Coleman, J.A. Durant, and L.A. Newman. 2006. Survival and growth of 31 Populus clones in South Carolina. Biomass and Bioenergy 30:750-758.        [ Links ]

Degen, A.A., K. Becker, H.P.S. Makkar, and N. Borowi. 1995. Acacia saligna as a fodder tree for desert livestock and the interaction of its tannins with fiber fractions. Journal of the Science of Food and Agriculture 68:65-71.        [ Links ]

George, N., M. Byrne, B. Maslin, and G. Yan. 2006. Genetic differentiation among morphological variants of Acacia saligna (Mimosaceae). Tree Genetics and Genomes 2:109-119.        [ Links ]

Hnatiuk, R.J., and B.R. Maslin. 1988. Phytogeography of Acacia in Australia in relation to climate and species-richness. Australian Journal of Botany 36:361-383.        [ Links ]

Holmgren, M., B.C. López, J.R. Gutiérrez, and F. Squeo. 2006. Herbivory and plant growth rate determine the success of El Niño Southern Oscillation-driven tree establishment in semiarid South America. Global Change Biology 12:2263-2271.        [ Links ]

Howard, D., G. Krebs, and M. Van Houtert. 2002. The value of Acacia saligna as a source of feed for sheep. Conservation Science of Western Australia 4:135-138.        [ Links ]

Jara, P.A., G. Arancio, R. Moreno, and M.R. Carmona. 2006. Factores abióticos que influencian la germinación de seis especies herbáceas de la zona árida de Chile. Rev. Chil. Hist. Nat. 79:309-319.        [ Links ]

Mcdonald, M.W., and B.R Maslin. 2000. Taxonomic revision of the Salwoods: Acacia aulacocarpa Cunn. ex Benth. and its allies (Leguminosae: Mmosoideae: section Juliflorae). Australian Systematic Botany 13:21-78.        [ Links ]

Meneses, R.R., and HP. Flores. 1999. Evaluación de Acacia saligna como forraje suplementario de caprinos de reemplazo y adultos en el ultimo tercio de preñez y lactancia. Agricultura Técnica 59(1):26-34.        [ Links ]

Midgley, S.J., and J.W. Turnbull. 2003. Domestication and use of Australian acacias: case studies of five important species. Australian Systematic Botany 16:89-102.        [ Links ]

Montecinos, A., and R Aceituno. 2003. Seasonality of the ENSO-Related Rainfall Variability in Central Chile and Associated Circulation Anomalies. Journal of Climate 16:281-296.        [ Links ]

Mora, F., S. Perret, and R. Meneses. 2004. Selección de procedencias de Acacia saligna: clave para obtener más forraje. Tierra Adentro 54:38-40.        [ Links ]

Mora, F. and R. Meneses. 2004. Performance of Acacia saligna (Labill.) H. L. Wendl. provenances in the Coquimbo Region, Chile. Ciencia Florestal 14:103-109.        [ Links ]

Mora, F., and S. Perret. 2007. Aplicación de técnicas Bayesianas en el análisis genético de árboles forestales. Bosque 28:198-206.        [ Links ]

Mora, F., S. Perret, C.A. Scapim, E.N. Martins, and M.P Molina. 2007. Source-dependent blooming variability of Eucalyptus cladocalyx in the region of Coquimbo, Chile. Cien. Inv. Agr. 34:99-106.        [ Links ]

Mora, F., R. Gleadow, S. Perret, and C.A. Scapim. 2009. Genetic variation for early flowering, survival and growth in sugar gum (Eucalyptus cladocalyx F. Muell) in southern Atacama Desert. Euphytica 169:335-344.        [ Links ]

Nativ, R., J.E. Ephrath, PR. Berliner, and Y. Saranga. 1999. Drought resistance and water use efficiency in Acacia saligna. Australian Journal of Botany 47:577-586.        [ Links ]

Ormazábal, C.S. 1991. Silvopastoral systems in arid and semiarid zones of northern Chile. Agroforestry Systems 14:207-217.        [ Links ]

Otieno, D.O., J.I. Kinyamario, and T.O. Omenda. 2001. Growth features of Acacia tortilis ana Acacia xanthophloea seedlings and their response to cyclic soil drought stress. African Journal of Ecology 39:126-132.        [ Links ]

Rodovalho, M.A., F Mora, E.M. Santos, C.A. Scapim, and E. Arnhold. 2008. Survival heritability in 169 families of white grain popcorn: a Bayesian approach. Cien. Inv. Agr. 35(3): 255-260.        [ Links ]

SAS-Institute. 1996. Statistical analysis system: user's guide. SAS Institute, Cary.        [ Links ]

Schneider, P.R., F.O. Fortes, L.H.S. Souza, A.D. Lucio, CA.G Finger, and PS.P Schneider. 2005. Análise da mortalidade de Acacia mearnsii De Wild. Ciencia Florestal 15:137-143.        [ Links ]

Squeo, F.A., Y. Tracol, D. López, J.R. Gutiérrez, A.M. Cordova, and J.R. Ehleringer. 2006. ENSO effects onprimary productivity in Southern Atacama desert. Advances in Geosciences 6:273-277.        [ Links ]

Sykes, R., B. Li, F. Isik, J. Kadla, and H.M. Chang. 2006. Genetic variation and genotype by environment interactions of juvenile wood chemical properties in Pinus taeda L. Ann. For. Sci. 63:897-904.        [ Links ]

Tiedeman, JA., and D.E. Johnson. 1992. Acacia cyanophylla for forage and fuelwood in North Africa. Agroforestry Systems 17:169-180.        [ Links ]

Van Tassell, C.P, and L.D. Van Vleck. 1996. Multiple-trait Gibbs sampler for animal models: flexible programs for Bayesian and likelihood-based (co)variance component inference. Journal of Animal Science 74:2586-2597.        [ Links ]

Velarde, M., P Felker, and C. Degano. 2003. Evaluation of Argentine and Peruvian Prosopis germplasm for growth at seawater salinities. Journal of Arid Environments 55:515-531.        [ Links ]

Wenninger, E.J., and R.S. Inouye. 2008. Insect community response to plant diversity and productivity in a sagebrush-steppe ecosystem. Journal of Arid Environments 72:24-33.        [ Links ]


Received June 03,2009. Accepted September 07,2009.

Corresponding authors:

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons