Fennel (Foeniculum vulgare Mill.) yield prediction using a regression model

Abstract

Fennel (Foeniculum vulgare Mill.) belonging to the Apiaceae family is a valuable medicinal, spice, aromatic, melliferous, vegetable and ornamental plant. A successful cultivation of fennel in the southern Steppe of Ukraine requires scientific modelling and prediction of the seed yield, taking into account different parameters of the technological practices and their effect on the productive processes of plants. The field experiments were carried out in 2011-2013 in the Kherson region on the dark chestnut soils typical for the southern Steppe of Ukraine. The experimental design included the following factors and their variants: Factor A – nutrition background: without fertilizers; N30; N60; N90; Factor B – sowing date: early (the third ten-day period of March at the right soil tilth stage); mid-time (the first ten-day period of April); late (the second ten-day period of April); Factor C – row spacing, cm: 15; 30; 45; 60. The model testing was conducted in 2014-2019 in the farms of Kherson region. According to the results of regression analysis the multiple linear regression model of fennel seed productivity depending on the studied technological practices was created: Y = 7.2302 + 0.0024Х1 – 0.0039Х2 + 0.0012Х3, where Y – fennel seed yield, t/ha; X1 – nutrition background (dose of nitrogen fertilizers), kg reactant/ha; X2 – sowing date (the sum of effective temperatures above 10oC), oC; X3 – row spacing, cm. A comparison of the experimental yield data and the values of the yielding capacity, calculated using the regression equation, confirmed accuracy of the created multiple linear regression model of the fennel seed productivity. The range of variation in the residual seed yield, depending on the interaction of the investigated parameters of the technological practices, was -0.11…+0.16 t/ha. In most variants, difference indicator between the true and modelled values of the crop productivity was within limits of ±0.05 t/ha. The results of the developed model application in agricultural enterprises of the region confirm the relevance of a mathematical approach to increase the seed productivity of common fennel and reduction the potential yield losses. The introduction of statistical methods into the technology of fennel cultivation is a promising strategy for the improvement of the crop productivity.