Post-harvest soil nutrient prediction in hybrid castor (Ricinus communis l.) Cropping sequence using a multivariate analysis technique

Autor: R. Abishek, R. Santhi, S. Maragatham, S. R. Venkatachalam, D. Uma, A. Lakshmanan
Rok vydání: 2022
Předmět:
Zdroj: Journal of Applied and Natural Science. 14:946-953
ISSN: 2231-5209
0974-9411
Popis: In the era of precision agriculture, the fertilizer prescription based on the soil fertility status is much required. Analyzing the soil after each crop is necessary for fertilizer recommendation and developing an alternative technique to forecast the soil available nutrient value rather than analyzing the soil. Multiple linear regression (MLR) equation was developed using filed experiment data to predict the soil available nutrient in castor cropping sequence. The post-harvest soil available nutrient was considered as the dependent variable and the initially available soil nutrient values, fertilizer added, yield and nutrient uptake of castor as an independent variable. In general, the post-harvest soil nutrient model's prediction accuracy was notable and had a coefficient of determination of less than 0.90. By calculating the RMSE (root means square error), R2 value, the ratio performance to deviation (RPD) and, RE (relative error) the performance of the MLR model was confirmed.Using the validated model, post-harvest soil available nutrients were predicted and compared with laboratory tested soil available nutreints. It turned out that the established model is more precisely effective and equally precise. Fertilizer recommendation could be made to subsequent crop after hybrid castor using the predicted soil available nutrients.
Databáze: OpenAIRE