Predictive model for available soil phosphorus in Enamor Ener Woreda, southern Ethiopia

Abstract

AbstractBackgroundPredicting soil phosphorus (P) needs and P fertilizer requirements are important for plant nutrition and reducing environmental risk. Thus, the precise estimation of available soil P (ASP) is required for the better management of environmental resources and improving agricultural productivity. Yearly, direct measurements of most nutrients at large scales are time‐consuming and expensive within the economic point of view.AimTherefore, to tackle such limitations indirect measurement of them in an accurate estimating manner is a possibility.MethodsThus, during this study, an exponential regression model for predicting ASP was suggested, and it was estimated as a function of soil organic carbon (OC). OC and ASP were used to investigate and analyze the soil to determine empirical relationships between them. Determination of the mentioned nutrient within the laboratory is somewhat difficult and time‐consuming. Therefore, an exponential regression model for predicting ASP from soil OC was suggested, and it was estimated as a function of soil OC. The ASP predicted from the ASP‐OC model was compared to the ASP determined by laboratory test using the paired samples t‐test and the Bland–Altman approach using SPSS.ResultsThe ASP predicted by the ASP‐OC model was not significantly different from the ASP determined by laboratory test since (p > 0.05).ConclusionsThe findings of the study indicated that the ASP‐OC model provides a simple, economic, and brief methodology to estimate ASP, and so to predict ASP‐supported soil OC using the exponential regression model (Y = aebX), the suggested model based on the findings for ASP is ASP = 9.786 e0.486 OC with R2 = 0.901 often suggested.