SciELO - Scientific Electronic Library Online

vol.36 número1Controlled release formulations of herbicides based on micro-encapsulationEstimation of leaf area in pecan cultivars (Carya illinoinensis) 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.36 n.1 Santiago abr. 2009 

Cien. Inv. Agr. 35(1): 43-52. 2009



Estimation of soil erosionability in the stream basin of Pillahuinco Grande, Province of Buenos Aires, Argentina


Fernanda J. Gaspari1, Alfonso M. Rodríguez Vagaría, and Gabriela E. Senisterra

Facultad de Ciencias Agrarias y Forestales, Universidad Nacional de La Plata. Diagonal 113 N° 469. CP 1900. La Plata. Provincia de Buenos Aires, Argentina.


In a hydrographic watershed, the erosionability (K factor) of the soil evaluation depends on the technical resources, developed applied sciences, and spatial technology available. For the determination of erosionability in the watershed of Pillahuinco Grande s Creek (Argentina) (38°S, 61°15'W), the Universal Soil Loss Equation (USLE) was used by applying a geographic information system (GIS) for the cartographic evaluation. A database of geological, environmental, and soil associations was developed, which indicated that geomorphologic variability, caused by geology and soil, significantly determines the spatial variation of the K factor values, in a range from 0.02 to 0.69 ((t·m2·h)·(ha·J·cm)-1). A new quantification was determined with the simplified K factor from the USLE model and with the two generated equations, K1 , starting from sand, silt and organic matter, and K2, beginning from sand and organic matter. A linear regression and the coefficient of efficiency (RN2) were established, which indicated the adjustment of the K factor for each developed pattern. The RN2 correlation value was 0.76 in relation to the simplified K factor, 0.87 to K1 , and 0.86 to K2 (very similar due to the small significance of silt in the equation). This relationship demonstrates that the application of K1 and K2, according to the readiness of the data, is more specific and exact than the results obtained by applying the simplified K factor.

Key words: GIS, soil, soil erosion , USLE, watershed.


La evaluación de la erosiónabilidad de los suelos (K) en una cuenca hidrográfica, depende de la disponibilidad de los recursos técnicos, ciencias aplicadas desarrolladas y de tecnología espacial. La metodología utilizada para su determinación en la Cuenca del Arroyo Pillahuinco Grande (Argentina), fue la Ecuación Universal de Pérdida de Suelo (USLE), establecida a partir de la evaluación cartográfica con aplicación de sistema de información geográfica. Se desarrolló una base de datos de ambientes geológicos y asociaciones de suelos donde se observó que la variabilidad geomorfológica, causada por la geología y suelos, determina de manera significativa una variación espacial de los valores del factor K, encontrándose en un rango de 0,02 a 0,69 ·m2-h)-(ha·J·cm)-1. A partir de K se estableció una nueva cuantificación con el K simplificado de la USLE y con dos ecuaciones generadas K1 a partir de arena, limo y materia orgánica, y K2 con arena y materia orgánica. Se estableció una regresion lineal y el coeficiente de eficiencia (RN2) que indicó el ajuste de K para cada modelo desarrollado. Este último, expresó una correlación de 0,76 en relación con K simplificado, de 0,87 con K1 y de 0,86 con K2, muy semejantes debido a la baja significancia del limo sobre la ecuación. Esta relación demuestra que la aplicación de la K1 y K2, según disponibilidad de datos, es mas precisa y exacta que los resultados que pude aportar la fórmula del Ks.

Palabras clave: Cuenca, erosionabilidad de suelos, SIG, suelos, USLE.


The hydrographic basin is the unit of study and admimstration in agrohydrological management, and it is defined as a territorial space formed by a mam water stream, its tributarles, and a water collector area, which are separated by the water dividing line. 

The presence of man in a basin creates a series of problems that arise from an irrational use of natural resources. Among these are the degradation and loss of soil productivity, erosion , floods, desertmcation, water eutrophicationforest destruction and loss of biodiversity. Because of these problems, the human population experiences a decrease in quality of life, which in extreme cases, ends with migrations towards large cities (Gaspari, 2000).

The planning and management of sustainable  basin development are necessary for understanding and stabilizing the use and management of natural resources and as a way to establish a methodology for planning the territorial arrangement of the Buenos Aires mountainous basins (Gaspari, 2000).

The hydrological restoration of forest, the active response of man to the destruction or deterioration of soil resources, vegetation and water, is the due to actions that must be addressed globally, which sometimes entail high, continuous investments for long periods. Their effects are beneficial only through time (Del Palacio, 1999).

In the mountainous basins of the Coronel Pringles Partido, Province of Buenos Aires, economical growth is historically related to agricultural and livestock production. The agricultural production corresponds to 50% of the district's production. The main crops are wheat, soybean, barley, sunflower, sorghum and maize, and 70% of the production is exported. Coronel Pringles is currently part of Consorcio Intermunicipal de Desarrollo Regional (CIDERE) that recent-ly opened the Regional Development Agency (Agencia de Desarrollo Regional, ADR). It is intended to foster strategic sectorial alliances, coordínate public and private efforts for fostering regional development, and promote cooperation, employment generation, and economical, social, and environmental sustainability. The area of study comprises the Pillahuinco Grande stream basin located in the Mountainous System of Ventania Partido of Coronel Pringles, in the southwest of the Buenos Aires Province, Argentina. It covers a surface of 109,353,95 ha (Figure 1). Serious environmental problems are observed in this zone, due to the degradation and loss of the soil surface by water erosion (Gaspari and Rodríguez Vagaría, 2006).

The surface water erosion results in the loss of the productive soil potential. Therefore, it is important to study the processes generating these problems. There are several factors affecting soil stability, which can be grouped as climatic, edafic, terrain or vegetation factors (Vich, 1989; Kirkby andMorgan, 1994).

Numerous methods have been developed for determining soil losses. In 1965, Wischmeier and Smith presented a model called the USLE (Universal Soil Loss Equation), in which the factors determining the amount of aroused sediments include a fifth factor of erosion control by crop practices. It is expressed asA = RxKxTSxC x P, where A is the soil losses (t·ha1year1), R is the index of pluvial erosion ((J·cm)·(m2·h)-1), K is the soil erosionability factor ((t·m2·h)·(ha·J·cm)-1), LS is the pending factor, C is the crop factor, and P is the dimensionless crop practices factor (Mintegui Aguirre and Tópez Unzu, 1990; Kirkby and Morgan, 1994; Tópez Cadenas de Llano, 1998; Mintegui Aguirre et ai, 2006).

Regardless of other factors of erosion , the susceptibility of soils to water erosion is determined from the content , texture , and structure of organic maller (OM), which are unlikely lo be available for a whole hydrographic basin (Irurlia et al, 1984).

The objeclive of this work was lo eslimale the coefficienl of soil erosionability (K) adjusted lo the Pillahuinco Grande slream basin according lo geospatial distribution. The zoning and adjustment of K will generale a dynamic digilal tool for determining the soil losses by surface water erosion in the Pillahuinco Grande stream basin.

Materials and methods

The stream basin of Pillahuinco Grande (38°S, 61°15'W) is characterized by a highly variable, temperate, and sub-humid climate with moderate temperatures.

Harrington (1947) described how the southern mountains of the Province of Buenos Aires form a cluster of elevations contrasting with the general level of the Pampas. The terrain presents different levels of erosion with drain lines of dendritic design, attached mainly to modern eolic sediments. The surrounding plain, free of rocky formations, presents a zone more immediate to the sierra (perimountainous) with uneven surfaces more or less noticeable, which are more evident near streams.

The geological formations present in the region were classified by Tricart (1973) as Platense eolic, Platense aqueous, Lujanian eolic and Lujanian aqueous, which explain the soil genesis of the area under study. Although all contribute to the soil formation in the region, the extensive mantle surface composed of the eolic post-Pampa sediments is the most important and pre-dominant material forming the soils.

The region under study offers a wide range of edaphic and environmental situations due to the diverse action from the factors mentioned above. From the topographical point of view, there are strong terrain variations that rule the arrangement and extension of consolidated rocky formations, overlaying sedimentary deposits, drain conditions, runoff surface and accumulation of rainwater participating in edaphic processes. The climate, terrain, structure and composition of the original geological material and the vegetation are factors that, in their regional variations, determine the edaphogenesis of the region's soils. The region is divided into four geological environments:

Mountainous environment. The extensive Lujanian eolic mantle along with the remaining Pampean tufa mantle and the Lujanian aqueous surfaces spreads over hillsides and summits and reduces the exposition of old consolidated rocks. Therefore, the rocky formations are scarce. The created terrain generally reflects the underlying rock, and the external form was adapted by an intense erosive action based on diverse geological structures. In the mountainous environment, the post-Pampean loess may be supported directly on old rocks.

Intramountainous environment. The intramountainous environment comprises the central and interior region of the sierra environment and includes the cluster of depressions and longitudinal and transversal valleys. The materials filling these depressions are predominantly thick and of the colluvial type that are in contact with mountains, of loessic character on the general plain, and are fluvial-lacustrian and partly covered by retransported edafic material towards the streams' axes. Because of its brittle nature, it is easily eroded by water action, which leads to the formation of large gullies. The streams cut the loessic mantle, and only the narrow fluvial bed maintains alluvial characteristics.

Perimountainous environment. This environment develops in attachment to the sierra and surrounds it and its slopes. When the transgressive accumulation of the Lujanian eolic deposits occurs, an inclined plañe is defined, leaning against the mountains, and a slope decreases slowly to fuse with the plain. The dense loess accumulation in the foothill slowly decreases and is thus present only at isolated spots. The most common places are the hills, which originated by the dissection of the general plain produced by streams running downthe mountains.

Plain environment. This environment extends externally to the perimountainous environment and is characterized by the loss of loessic cover homogeneity as the presence of tufa becomes more abundant. However, some aspects of the plain are very different, so different sectors are noteworthy: a. Northern Sector: Inthe extended plain sector, it presents homogeneous characteristics. The fluvial streams extend in scarcely carved wider valleys subject to flooding. The unevenness between the stream beds and the plain reaches a barely noticeable inflexión, b. Spills Sector: The mountainous streams spread superficially on the plain to generate "spills." The detritic material is characterized by abundant dissolved calcium carbonate, which forms a calcareous layer through evaporation. During the last dry period, the floods concéntrate the majorproportionof soluble salts (sodium), causing salination and/or alkalization of the affected areas, c. Southeast Sector: The Lujanian eolic accumulation is dense and abundant at the foot of the mountains and becomes thinner farther away from the mountains. The accumulation re-speets the pre-existing terrain and maintains the topographical relationships.

Soil identification (Table 1), geological associations, distributions and local variations allow for the classification of the environment. As the soil is an integral and functional part of the environment where it is present, it is important to show how geomorphologic variations influence soil constitution and how the edaphological classification adjusts to the surrounding geological environments (Table 2).

The Geographic Information System (Sistema de Información Geográfica, SIG) allowed for the elaboration of the edaphic base cartography and geological associations of the basin of the Arroyo Pillahuinco Grande (Idrisi Kilimanjaro 14.0, Clark University, USA, 2003). The procedure consisted of generating and zoning the basin according to the maps of soil association and geological environments, by means of digitaliing the analogical format (scale 1:250.000).

In orderto determine the soil erosionability (K), the only published data (Spinelli Zinni et al, 1978) were used to elaborate, arrange, analyze, classify and interpret a geospatial database.

The development and application of a cartographic model allowed for data analysis and re-cording with the SIG Idrisi Kilimanjaro, which permitted space tracking (López Cadenas de Llano, 1998; Gaspari et al, 2006).

The texture, OM, permeability and structure data were extracted from the most representative area profiles according to Spinelli Zinni (1978). In order to determine the factor K, it was necessary to identify the representative profile of the series type (Scotta et al, 1986). It is worth mentioning that these profiles were extrapolated according to the relationship of soil association and location in respect to the geological environments.

The soil erosionability (K) (t·m2·h)·(ha·J·cm)-1 was calculated by Formula 1, according to the universal equation of soil loss (USLE) (Mintegui Aguirre and López Unzú, 1990):

where M is the product of soil particles between 0.002 and 0.1 mm of diameter and is expressed in percentages by the percentage of soil particles between 0.002 and 2 mm in diameter; a is the percentage of OM (for the calculations, 1.72 of organic carbon was used when the OM percentages were not available); b represents the soil structure coded according to the granule size; c describes the class of permeability soil, according to the USDA (Wischmeier and Smith, 1978).

For the estimation of simplified factor K (Ks), the values b and c may be dispensed, and Eq 2 is obtained (Mintegui Aguirre and López Unzú. 1990):

Once the erosionability K factor and Ks for each profile were determined, an area weighted by the factor () was made according to the area of influence to establish the coefficient for each soil association according to the geological environment.

These results allowed for the generation of a mathematical adjustment model adapted to the Pillahuinco Grande stream basin, producing two equations to modify and promote the K determination factor. Thus, laboratory data obtained by processing simple soil samples taken from field were used and corroborated with data records from studies by Spinelli Zinni (1978). From the development of a regression equation and the inclusion of the percentages of sand, silt and OM, the K1 value was obtained. A second regression equation showing how the independent variables, sand and OM, allowed for obtaining the K2 value. These two models allowed for representing the soil erosionability in the basin in a simplified and practical form.

The goodness of fit between the data obtained with the original model (Eq 1) and the data obtained with the other three models (Ks, K1 and K2) was evaluated by a linear regression analy sis with a coefficient of determination (R2) by the minimun square method (Navidi, 2006) and the criterion of efficiency by Nash-Sutcliffe (1970). The Nash-Sutcliffe regression (RN2) represents the relationship between the observed and predicted values (Llorens, 2003).

Results and discussion

The zoning of the geological environments and the soil associations are presented in Figure 2A and B, respectively. The contributions of dif-ferent soil associations with respect to the geological environments in the Pillahuinco Grande stream basin are shown in Table 2.

The database, derived from the study by Spinel-li Zinni (1978) and including the soil profile according to the geological environment, is pre-sented in Table 3. The factor , representing the occupation degree, was used to weight the K value in homogeneous surfaces for each soil association by implementing the methodology in Eqs 1 and 2. The weighted K values and simplified K values varied according to the intrinsic characteristics of each homogeneous zone determined by SIG, where the content of OM was the most influential variable in the minimum extreme of K and Ks values, and the soil structure and permeability were most influential in the máximum extreme. This presented a range between 0.0144 and 0.6874 (t·m2·h)·(ha·J·cm)-1 for K and a narrower range, between 0.0467 and 0.5121 (t·m2·h)·(ha·J·cm)-1, for Ks.

The mathematical K model adjusted to the zone under study established two new equations. The first (Eq 3) was determined through a linear múltiple regression, in which the percentages of sand, silt and OM were used as independent variables and the factor K calculated from the USLE model was a dependent variable:

According to the R2 value , 87.35% of the variability of factor K1 is explained by the three variables (sand, silt and OM), with a 99% con-fidence level. By individually analyzing each variable, the value of probability for silt is 0.49, which is non-significant and with a confidence level of 90%. The analysis of the independent variables of K1 determined that the model may be simplified to a new expression, given by Eq 4 (K2):

The R2 value for K2 explains that 86.76% of the variation was due to the sand and OM values. In order to establish the relationships among the formulas, a simple linear regression was used (R2), where the K was correlated with Ks, K1 and K2 (Figure 3A, B and C). The dispersion of the data expressed with the graphics allowed for verifying the goodness of fit among the different K values. This was expressed with a major adjustment of R2 with the relationship between K and K1 (0.86), without significara differences with the K and K2 (0.85) relationship. In the case of Ks, an R2 value of 0.76 was obtained.

The application of RN2 expressed the results homologically, since a value of 0.87 with a K2 of 0.86 was obtained in the relationship of K with K1 . However, the value with Ks was 0.67.

The regression analysis and the use of the criteria by Nash-Sutcliffe (1970) established that the proposed formulas adjust more precisely with the K value from the USLE than the simplified formula.

The obtained results allowed for defining a new methodology for the determination of soil erosionability in the stream basin of Pillahuinco Grande according to geospatial distribution This erosionability was established simply with a high correlation to the values generated by K from the USLE and without the need of a wide database. Furthermore, the sand and OM data are inexpensive and easy to access and determine.

The presera work includes new elements for the identification of the physical state of a hydrographic basin using a universal basic methodology and SIG. Although many investigators criticize the USLE due to a weak scientific base, it has been an effective tool to predict erosion and design strategies for soil conservation (Kirkby and Morgan, 1994).

The present work allowed for the estimation of the coefficient of erosionability and its space distribution for the Pillahuinco Grande stream basin, Argentina, through four parametric models, based on the factor K from the USLE, which is of great importance for evaluating the risks of surface water erosion . The geomorphologic variability, caused by geology and soils, significantly affected the spatial variation of the factor K. The zoning and adjustment of K generated a dynamic digital tool for using the USLE to determine erosionability for the estimation of soil loss by surface water erosion .


This work was carried out with financing from Proyectos de Investigación Científica y Tecnológica Orientados, PICTO 2003 N°07-13.741, Secretaria de Ciencia y Técnica (SE-CyT), Argentina.



Del Palacio, E. 1999. La restauración hidrológica forestal en España. Gestión sostenible de los recursos suelo, agua y vegetación. Ministerio de Medio Ambiente. Madrid, España. 75 pp.        [ Links ]

Gaspari, F.J. 2000. Ordenamiento Territorial en Cuencas Serranas Aplicación de Sistemas de Información Geográfica (SIG). Ediciones Cooperativas. Buenos Aires, Argentina. 116 pp.        [ Links ]

Gaspari, F.J., and A. Rodríguez Vagaría. 2006. Zonificación Ambiental de la Cuenca Pillahuinco Grande. Pages 9-10. In: XI Reunión Argentina de Agrometeorología. Facultad de Ciencias Agrarias y Forestales. Universidad Nacional de La Plata. Argentina.        [ Links ]

Gaspari, F.J, M.G., Leonart, C. De La Peña, and A. Rodríguez Vagaría. 2006. Cartografía temática para la evaluación de la erosión hídrica superficial de la Cuenca del Arroyo Pillahuinco Grande. Provincia de Buenos Aires. Argentina. Pages 220-233. In: Tercer Congreso de la Ciencia Cartográfica. X Semana Nacional de la Cartografía. Centro Argentino de Cartografía. Buenos Aires. Argentina.        [ Links ]

Harrington, H. 1947. Explicación de las cartas geológicas 33 m (Sierra de Cura Malal) y 34 m (Sierra de la Ventana), Provincia de Buenos Aires. Dirección de Minería y Geología. Volumen 61. Buenos Aires, Argentina.        [ Links ]

Irurtia, C, J. Musto, and P Culot. 1984. Evaluación de Riesgo de Erosion Hídrica en el Sector Argentino de la Cuenca del Plata. Publicación N° 174, Instituto Nacional de Tecnología Agropecuaria (TNTA). Castelar, Buenos Aires, Argentina. 32 pp.        [ Links ]

Kirkby, M.J., and R.P.C. Morgan. 1994. Erosion de Suelos. Editorial Limusa. Noriega Editores. México, DF. México. 375 pp.        [ Links ]

López Cadenas de Llano, F. 1998. Restauración Hidrológica Forestal de Cuencas y Control de la erosión . Editorial TRAGSA. Madrid. España. 945 pp.        [ Links ]

Llorens, P 2003. La evaluación y modelización del balance hidrológico a escala de cuenca. Ecosistemas 2003/1. (Accessed: January 2007).        [ Links ]

Mintegui Aguirre, J. A., and F. López Unzú. 1990. La Ordenación Agrohidrológica en la Planificación. Servicio Central de Publicaciones del Gobierno Vasco. Madrid. España. 308 pp.        [ Links ]

Mintegui Aguirre, JA., J.C. Robredo Sánchez, J.I. García Viñas, and C. López Leiva. 2006. Introducción a la restauración hidrológicos-forestal de cuencas hidrográficas. Rev. Ecología 20:389-414.        [ Links ]

Nash, JE., and IV Sutcliffe. 1970. River flow forecasting through conceptual models. A discussion of principies. Journal of Hydrology 10:282-290.        [ Links ]

Navidi, W. 2006. Estadística para ingenieros y científicos. McGraw Hill Interamericana. México DF. México. 868 pp.        [ Links ]

Scotta, E.S., LA. Nani, AA. Conde, A. Rojas, H. Castañeira, and O. Paparotti. 1986. Manual de Sistematización de Tierras para Control de Erosion Hídrica y Aguas Superficiales Excedentes. TNTA, Serie N° 17. Castelar. Argentina. 51 pp.        [ Links ]

Spinelli Zinni, F. 1978. Estudio de situación Partido de Coronel Pringles, Provincia de Buenos Aires. INTA. Tomo III. Buenos Aires. Argentina. 92 pp.        [ Links ]

Tricart, J.L. 1973. Geomorfología de la Pampa Deprimida. Instituto Nacional de Tecnología Agropecuaria (INTA). Colección Científica XII. Buenos Aires, Argentina. 202 pp.        [ Links ]

Vich, A.I.J. 1989. Erosion hídrica: Estimación y medición de pérdidas de suelo. Pages 118-130. In: Curso Latinoamericano sobre detección y control de la desertización. Editorial FA. Roig. CRICYTME. Buenos Aires. Argentina.         [ Links ]

Wischmeier, W.H., and D.D. Smith. 1978. Predicting rainfall erosion losses. A guide to conservation planning. U.S. Department of Agriculture. Agriculture Handbook N° 537.58 pp.        [ Links ]


Received 05 March 2007. Accepted 25 September 2007.

1 Corresponding author:

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