Modeling a Repeatedly Measured Farmyard Manure and Gypsum Intervention on Yield of Sorghum in Saline Sodic Soils

Abstract

Background: Salinity is one of the major environmental problems the world is currently dealing with, and it also poses the biggest obstacle to agricultural productivity, especially in arid, semiarid, and dry sub humid regions. Sorghum is the most produced grain in Raya Alamata, Southern Tigray of Ethiopia. Soil salinity in this area often affects sorghum yield and yield characteristics. In order to tackle the effect soil salinity/sodicity status, ongoing assessment and monitoring on fertilizer intervention should be conducted in the stated study area Objective: The aim of this thesis is, hence, twofold: to propose repeated measure analysis to take into account the problems with traditional approach; and make use of this approach to scrutinize the optimum rate of gypsum and farmyard manure fertilizers for improving sorghum yield in saline sodic soils of Raya Alamata, Southern Tigray of Ethiopia with an emphasis on the use of mixed modeling techniques with repeated measures. Method: The study involved secondary repeated measure or longitudinal data, collected using field experiments during three consecutive years of the same plots to illustrate different modeling strategies and graphical tools with an emphasis on the use of mixed modeling techniques with repeated measures. Mixed modeling approach is the most flexible method in terms of handling the covariance among repeated measures. Result: The Breusch Pagan method shows Pooled OLS model is not the best estimation method (p-value=0.000). Hausman Test output also indicates random effect model outperforms compared to the fixed effect (p-value=1.000). The spaghetti plot showed that the increasing trend with time. A positive linear relationship was also observed between yield measurements taken at different years, confirming the strong correlation. The mean yield with 0% GR was 34.23qt/ha but 36.33 qt/ha and 37.58 qt/ha after addition of Gypsum 50%GR and 100% GR respectively. Similarly there was a better yield after adding FYM when compare to the control. The mean yield is increasing over time, the mean yield is 35.47, 35.94 and 36.72 at first, second and third year respectively. The maximum yield 40.25 also recorded in the third year. Sole application of 8 tha-1 FYM and combination of FYM and Gypsum rate at (50%GR+8 tha-1 FYM) have significance role on production yield of sorghum on salinity area of Raya Alamata over three consecutive years with p-value < 0.000 at =5% level of significance respectively. Conclusion and Recommendation: The study found that repeated measure analysis using mixed modeling techniques yields a better result than a traditional approach (OLS). The result from the Bonferroni mean comparison further revealed that combination 8tha-1 FYM and 50% GR are the optimum level of fertilizers that provide improved sorghum yield in saline sodic soils of Raya Alamata, Southern Tigray of Ethiopia.