Fig 1.
Map of Iran with the main provinces where olive accessions had been sampled.
Blue) local ecotypes; green) cuspidata specimens.
Fig 2.
Examples of indigenous Iranian olive.
A) Torang cuspidata specimen, Kerman; B) Mavi local ecotype, Khuzestan; C) Gardineko local ecotype, Ilam; D) Pirzeytun local ecotype, Fars; adapted from Hosseini-Mazinani et al. [36].
Fig 3.
Decision Tree generated model showing separation of olive populations in the 4-targeted (4-t) experiment by different alleles.
In this model, DCA14-149 was selected as the root.
Fig 4.
Decision Tree generated model showing separation of olive populations in the 16-targeted (16-t) experiment by different alleles.
In this model, DCA-178 was selected as the main classifying attribute.
Table 1.
Microsatellite allele lengths, loci and the total alleles.
Table 2.
Prediction rate (accuracy) details of each decision tree with 10-fold cross validation for each of the populations in the 4-targeted (4-t) experiment, i.e. reference cultivars, Mediterranean varieties, cuspidata specimens, and local ecotypes.
Table 3.
Prediction rate (accuracy) details of each decision tree with 10-fold cross validation for each of the types in the 16-targeted (16-t) experiment.
Table 4.
Prediction rate (accuracy) details of each Bayesian algorithm with 10-fold cross validation for each of the populations in the 4-targeted (4-t) experiment, i.e. reference cultivars, Mediterranean varieties, cuspidata specimens, and local ecotypes.
Table 5.
Prediction rate (accuracy) details of each Bayesian algorithm with 10-fold cross validation for each of the populations in the 16-targeted (16-t) experiment.