Path-coefficient and stepwise regression analyses under different sowing dates in alfalfa (Medicago sativa).

Abstract

The current work was carried out at Agronomy Department Farm, Faculty of Agriculture, Assiut University, Egypt, to study the direct and indirect effects and stepwise regression of yield attributes on forage and seed yields of ten alfalfa genotypes. Two experiments were carried out as experiment I (2017-2019) and experiment II (2018II ( -2020. The experiments were laid out in a split block design with three replications. Five sowing dates at 10 th October (D1), 10 th November (D2), 10 th December (D3), 20 th March (D4) and 20 th April (D5) were used. Fifteen cuts were taken from each sowing date in two years, while, seed yield was taken in the second year which take in the first week of March, April and May for the studied sowing date. The obtained results of path analysis, show that, fresh forage yield has the greatest influence on protein forage yield in each sowing date. However, its indirect effects were negative via mean dry matter percentage or negligible via protein percentage in both experiments. Also, the number of pods/plant and number of seeds/pod contributed to most direct effect on seed yield/plant for each sowing dates. Meantime, the results of stepwise regression analysis showed that, the model no.2 which included two independent traits i.e. dry forage yield and protein percentage resembled the fit model for protein yield. Also, the model no.2 which possessed two independent traits i.e. number of seeds/plant and 1000-seed weight has strong contribution to seed yield/plant.