Evaluation of models for estimating electrical conductivity and potassium concentration of the soil solution by using time domain reflectometry (TDR) / Avaliação de modelos para estimar a condutividade elétrica e a concentração de potássio na solução do solo usando Reflectometria no Domínio do Tempo (TDR)

Gessionei da Silva Santana

Time domain reflectometry is a technique that has been outstanding in the last years as an attractive tool for soil water content (θ) and for bulk electrical conductivity (ECa) monitoring on real time, automatically, with the minimum soil disturbance. Besides, this technique allows the knowledge of electrical conductivity (ECw) and nutrient concentrations (Ci) of soil solution, indirectly (WRAITH &DAS, 1998; MMOLAWA &OR, 2000 e NOBORIO, 2001). This work had as objective to evaluate six models that relate θ, ECa and ECw (RHOADES et al., 1976; NADLER et al., 1984, modified by RHOADES et al., 1989; RHOADES et al., 1989; MUALEN &FRIEDMAN, 1991; HEIMOVAARA et al., 1995; VOGELER et al., 1996), concerning its capability of estimating the ECw and the potassium concentration in the soil solution from data of θ and ECa, obtained by means of TDR technique as well to evaluate the feasibility of using TDR technique for monitoring time and space variation of θ, ECa and ECw, under field conditions. In order to estimate K, the models were adapted with potential (MMOLAWA &OR, 2000) and linear (HEIMOVAARA et al., 1995; VOGELER et al., 1996) functions relating ECw and K. During the laboratory phase, alluvial soil classified as loam (CTf) and sand loam (CTfa) was packed in pots in order to get the same soil densities from field. Five potassium chloride solutions of electrical conductivities of 1,0; 2,5; 4,0; 5,5 and 7,0 dS m-1 were applied on 25 pots with each type of soil so that five soil water contents as 20; 40; 60; 80 and 100% of the available water could be obtained. During the field phase, two trenches were dug in a banana orchard that used to be fertirrigated by a microsprinkler irrigation system. The trenches were opened in the direction pseudo stem - emitter and TDR probes were installed in their walls. The trenches were filled and no readings were made for 60 days. The banana crop was fertilized with 432 kg ha-1 of K2O per year in one trench and in another it was fertilized with 1,008 kg ha-1 of K2O per year. For both phases (laboratory and field) readings of θ and ECa were done by means of TDR equipment. The soil solution was extracted with water samplers in order to determine ECw and K. The ECw was determined by using a desk conductivimeter and potassium was determined by a flame spectrophotometer. The monitoring of θ and ECa was done in 22 locations of the soil profiles during field phase. TDR equipment was connected to four multiplexers with eight channels each where 22 TDR probes were plugged. ECw was also monitored in these same locations. In each trench, θ, ECa and ECw were read during eight days for two fertirrigation events. The readings were performed in a 15-minute intervals and stored in a datalogger. Soil solution was collected 45 minutes before and after each fertirrigation event and in intervals of 24 hours until 24 hours after the second fertirrigation event. After have fitted the models to data of θ, ECa and ECw and have determined the model parameters, CEw was estimated for the best model in the whole profile monitored with the TDR (22 locations). Afterwards, profiles of θ, ECa and CEw were made for the following moments: three hours before three hours after each fertirrigation event and from that time they were made in 24-hours intervals until 24 hours after the second fertirrigation event. The evaluation of models concerning their capability of relate θ, ECa and ECw and θ, ECa and K was accomplished based upon the agreement coefficient (D) proposed by WILLMONTT (1981), the goodness of fit (R2) and based upon the angular coefficient of the linear equation Y = aX. It s possible to estimate K and CEw from θ and ECa data, under laboratory conditions by using models of RHOADES et al. (1976), VOGELER et al. (1996) e MUALEN &FRIEDMAN (1991), adapted with a potential relation between in the range of 0 to 60 and 0 to 120 mg L-1, for CTf and CTfa soils, respectively. The models estimated reasonably CEw from θ and ECa obtained by TDR technique under field conditions, but this capability reduces with the increase of soil salinity, reducing data quality, as a consequence. TDR technique showed limitation in the monitoring of spatial and temporal variation of ECa and ECw, under field conditions, mainly in case higher amount of K2O applied (1,008 kg ha-1).

AUTOR(ES)

DATA DE PUBLICAÇÃO

RESUMO

ASSUNTO(S)

ACESSO AO ARTIGO

Documentos Relacionados