Applicability of Regression Models Between Yield Parameters and Some Soil Properties of Maize Plant

The aim of this study was to set regression models between yield parameters of maize plant grown in Çarşamba Plain (plant height, thousand seed weight and grain yield) and some physical and chemical properties of soils and to determine applicability of obtained models in estimation of yield in plain soils. For this purpose, soil and plant samples were taken from root zone of cultivated farms. In the research area, 65% of the maize plant height values were determined as between 250 cm and 300 cm, 60% of thousand seed weight values were between 300 g and 460 g, and 57.5% of the yield values were between 1000-1400 kg da-1. In the regression model between maize plant height and electrical conductivity (EC), Ca+Mg, lime (CaCO3 ), bulk density (BD), field capacity (FC), organic matter (OM), (EC)2 , √Sand, nitrogen (√N) content wasn’t statistically significant (R=0.543, p>005); in the model between 1000 seed weight of maize and (EC), (OM), Sand, (BD)2 , (Sand×BD), √(CaCO3 ), √EC , wilting point (WP) parameters, the high determination coefficient R = 0.819 was determined at statistical significance level of p = 0.012; the performance of the model between maize grain yield and OM, N, phosphorus (P), potassium (K), sodium (Na), zinc (Zn), Ca+Mg, Silt, Sand, (BD×WP), BD, WP parameters was very high (R = 0.840; p = 0.001). Determination coefficient (R), root mean square error (RMSE), index of agreement (d), model efficiency (ME) were evaluated together to determine the validity of regression models between the yield components and physical and chemical properties of soils. In general, statistical parameters were within validity limits. It can be seen that the regression models obtained can be applied in the estimation of yield parameters of maize plant grown in plain soils.