Determination of the characteristics associated with salinity tolerance of wheat enotypes in Yazd province using logistic regression

Identification of pattern and classification is one of the most applications of statistical methods in the different sciences. One of the main objectives of modeling and classification is the predications based on variables and available data about a specific topic. In order to determination of the characteristics associated with salinity tolerance, 41 wheat genotypes were evaluated in two conditions (normal and stress) at the research field of the National Salinity Research Center (NSRC). The salinity of water used in irrigation in saline and non-saline conditions was 10 and 2 dS.m-1, respectively. In this study, in order to determination of the characteristics associated with salinity tolerance was used from logistic regression. The dependent variable in this research has the qualitative nature and scale value and other variables in this research were considered as independent variables. The results of the used parameters to goodness of fit logistic regression model indicated that logistic regression model is an appropriate model for determination of the characteristics associated with salinity tolerance in wheat genotypes. The results of logistic regression showed that five variable of number of fertile tillers, number of spikelets per spike, the 100-seed weight, plant height and peduncle length are the most important variables associated with salt tolerance. These variables have the high ability for identification of sensitive and tolerant genotypes under salinity stress in bread wheat. So that the results in study can be used in breeding programs to identify salinity tolerant genotypes.