Predictive model for fibre yield estimation of tossa jute (Corchorus olitorius) in India

Abstract

The present study explores the possibility of developing statistical models for prediction of dry fibre yield of tossa jute (Corchorus olitorius) in India. Sixty plants of 110 days age were randomly selected from the study area, and their basal diameter and dry fibre yield were measured for analysis and fitting of models. Basal diameter was found to be a very good predictor of dry fibre yield and was included in the model. Various functions (linear, monomolecular, logistic, Gompertz, allometric, Chapman) were fitted to estimate a relationship between dry fibre yield and basal diameter of the plant. The adjusted R2 values were > 0.72 for all six models and Chapman model (Y = k × (1-exp (-a×X))b) was found as best performing with highest R2 (0.77) lowest Akaike’s information criterion (AIC) value (68.40) . The paired t statistics value between observed and predicted fibre yield was found non-significant (p>0.05), which clearly indicates the validity of the model. The models developed in the present study will be of immense help to the jute cultivators in estimating dry fibre yield before harvesting, by simply measuring the basal diameter values of the jute plants. The estimated fibre yield is 34 q/ha with average basal diameter of 1.3 cm and 4 lakh/ha plant density.