Abstract Pyrus communis is a difficult-to-root species, and its optimal rooting depends on applying an accurate concentration of various chemicals. Since statistical modeling can be applied to estimate the approximate range of factors needed for the success of in vitro plant propagation, we tried to optimize the rooting medium using response surface methodology. Therefore, sucrose and polyvinyl alcohol were selected as independent variables in the central composite design of response surface methodology with α= 1. Based on the pre-test, it was found that the independent variables for the root formation were 2 to 3% sucrose and 0.5 to 1.5 g L −1 polyvinyl alcohol in studied cultivars. The statistical analysis results showed that the quadratic models were significant, with high R2 values > 0.9 for rooting rate response. The optimized values for independent variables in rooting of the studied cultivars included 2% sucrose, 1.5 g L−1 polyvinyl alcohol in ‘Williams’; 2.85% sucrose, 1.47 g L−1 polyvinyl alcohol in ‘Natanzi’; and 3% sucrose, 0.99 g L−1 polyvinyl alcohol in ‘Ghosi’. A separate validation experiment detected that the exper- imental values of the studied responses are in close agreement with predicted values in all cultivars. For the first time, the methodology was developed using a central composite design to optimize the rooting in P. communis

Keywords