A quantitative multivariate methodology for unsupervised class identification in pistachio (Pistacia vera L.) plant leaves size

Aim of study: Genetic diversity pistachio, can be evaluated by using different descriptors, as adopted in international certification systems. Mainly the descriptors are morphological traits leaf, which represents an important organ for its sensibility to growth conditions during expansion phase. This study a rapid and quantitative non-hierarchic clustering classification (k-means), extract size classes basing on contemporary combination (i.e., leaf stalk length, terminal width ratio) varietal collection composed 21 pistachio cultivars.Area Worldwide.Material methods: The unsupervised technique was entire samples leaves from k=2 k=15 both four variables three ratio).Main results: A model only (for results statistical analysis groups resulted more separated all variables), with k= 5 (five groups), constructed non-linear artificial neural network approach. percentages bad prediction training testing equal 0%. “terminal length” returned higher impact (44.89%).Research highlights: traits, allowed create automatic great importance cultivar identification comparison.