Servicios Personalizados
Revista
Articulo
Indicadores
Citado por SciELO
Links relacionados
Similares en SciELO
Compartir
Ciencia del suelo
versión On-line ISSN 1850-2067
Resumen
FAITA, ELIDA CRISTINA; ROS, MIGUEL ÁNGEL y GIORDANO BUIANI, ALICIA ROSARIO. FORMULACIÓN DE UN ÍNDICE DE CALIDAD EDÁFICA PARA ARGIUDOLES DE LA CUENCA MEDIA DEL RÍO LUJÁN. Cienc. suelo [online]. 2015, vol.33, n.1. ISSN 1850-2067.
In order to develop a soil quality index (ICE) applied to Argiudolls in the middle basin of Río Luján, six treatments were established at different sites: natural field with minimal disturbance (md), no-till (SD), no-till and deep plowing (SDS), reduced tillage and deep plowing (LRS), pasture (P) and conventional tillage with maximum disturbance (MD). The following soil properties were determined on three sites per treatment:: total organic matter (MO), microbial biomass (BM), bulk density (DAP), anisotropic index (IA), aggregate instability index (IS) and cumulative infiltration (INFAac). The following statistical analyses were performed on the soil data: analysis of variance (ANVA), principal component analysis (ACP) and canonical discriminant function (FDC). The following criteria was used to select soil quality indicators: statistical importance in explaining the variability, positive or negative correlation between variables and measuring complexity. When a performing an ANVA, the biological properties generated maximum significant differences among treatments. ACP eigenvalues and eigenvectors determined and assessed the relationships between properties and between properties and treatments. The eigenvalues of PC1 and PC2 (principal components) explained 59% of total variability. The eigenvectors expressed the highest positive correlations with PC1 for MO (0.57), BM (0.48), INFAC (0.48) and IA (0.41) and greater sensitivity to changes depending on use and handling. DAP and IS were dismissed due to low BM values for complex eigenvalues and measurement difficulty. MO (high positive correlation with BM), INFAC and IA were selected. When applying linear discriminant analysis (ADL) to the selected properties, the following index was obtained ICE = -8.08+2.27*MO+0.16*IA+0.01*INFac. ICE for each treatment determined a decreasing sequence linked to the degree of soil disturbance (md, P, SDS, SD, LRS, MD), consistent with available evidence and beyond differences within Argiudolls. The location of treatments for ACP in different quadrants generated three groups according to tillage intensity: SD and SDS, md and P, and MD and LRS, compatible with that generated by FDC.
Palabras clave : Soil quality, principal component analysis, canonical discriminant functions.