Variabilidade espacial dos atributos físico-hídricos e do carbono orgânico do solo de uma bacia hidrográfica de cabeceira em Canguçu - RS

Abstract

The population's demand for water is admittedly a factor linked to the quality of life and can be indicative of the economic and social development of a given region. Brazil, like other developing countries, has a shortage of information related to water resources, although it has a vast hydrological potential in most of its territory. In this context, the understanding of soil water dynamics is fundamental in the study of hydrographic basins, emphasizing mainly the influence on hydrological and hydrosedimentological variables. The monitoring of soil moisture in watersheds presents high complexity due mainly to its spatial and temporal variability. Soil hydro-physical attributes provide important information for several areas of knowledge, such as the estimation of soil susceptibility to erosion, hydrological modeling, irrigation projects, etc. The objective of this work was to characterize ten soil hydro-physical attributes in order to evaluate the spatial variability inherent to the physiographic heterogeneity of the landscape in a hydrographic basin and to validate the intrinsic hypothesis of geostatistics. The study was conducted in a sub-basin of Arroio Pelotas called the Sanga Ellert Basin (BHSE), located in the municipality of Canguçu. The basin has an area of approximately 70ha and the altitude varies from 310.9 to 419.4 meters. The studied areas basically comprise a class of soils, the Neosols. Descriptive statistics and geostatistics were used to determine the magnitude and spatial variability of the following soil physical-water attributes: Sand, clay, soil density, total porosity, macroporosity, microporosity, field capacity, permanent wilting point and saturated soil hydraulic conductivity and also to soil organic carbon,. The exploratory analysis of the data showed non-normality for most of the variables, except for the sand content, microporosity and field capacity. Falling on of classical statistics we can still emphasize expected correlations for these attributes demonstrated by the Pearson coefficient, as for the sand content and the field capacity; clay content and permanent wilting point; saturated soil hydraulic conductivity with soil density and macroporosity; and microporosity with field capacity. For all the attributes that presented non-normal series, the robust estimator of Cressie and Hawkins was used for the construction of semivariograms as well as for the attributes sand and field capacity, which presented outliers in their data. Only for the microporosity the Matheron's classical estimator has been used. The results showed that, except for the hydro-physical attributes sand and clay content, the other attributes evaluated in the present study demonstrated spatial continuity and that from the proposed sampling grid is possible to identify this condition, as well as to perform the data interpolation using the ordinary kriging technique.