Some remarks on variety trials in sugar beet

Accuracy of sugar beet trials, particularly variety comparisons, depends on several factors including; statistical design, treatments oriented, experimental techniques, and replication numbers in order to estimate various components of a statistical model precisely. Testing agronomic treatments or evaluating new varieties for plant protection (PVP) or value for cultivation and use (VCU) under both field and glass-house conditions needs a reliable statistical design. There are many factors and circumstances like response of genotype, location, year, agronomic techniques etc. that affect performance of a sugar beet trial. Some of these factors are usually unknown before conducting the experiment that may influence the precision and accuracy of a field experiment. In addition to these, uncontrolled variations are usually summed in the experimental error. Type and quantity of the “variance of error” may be originated from different sources in which mainly are the precision of measurements and techniques as well as choice of the experimental design. The experimental design is also function of a number of other factors. When using RCBD, plots should be oriented in such a way to minimize soil variation, nutrient and pesticide residuauls within a block. In practice, treatments should be applied in parallel to the directions of blocks of a trial to minimize variance of error. A well-prepared field trial is effective in reducing the experimental error; however, a good and homogenous field emergence depends on the quality of seedbed preparation. Size of plot and plant number within a plot is essentially important. Homogeneity of a plot depends on its size; too big plots increase the error and too small plots would leads to lack of precision. An optimum plot size and average number of plants at harvest time needs to be sufficient to stabilize the error at a low level. If a trial carried out at different locations for several years, effects of these factors must be considerd precisely. Random or fixed effects of entries are very important for the estimation of variances. In sugar beet variety trials, it is common to apply either RCBD or lattice designs, each of them has advantages and limitations. In order to reduce variance of error and consequently increasing experimental accuracy of variety comparison, complete block or incomplete blocks (lattice design) should be arranged on a homogenized soil. Since there are close competitions among commercial sugar beet varieties, an appropriate statistical model based on the type of variation sources is essential to select superior varieties within a trial and increases the efficiency of selection for superior lines and varieties in plant breeding programs.