Genome scan was done with 439 dominant marker loci to identify outlier loci in three populations of Morus alba from the trans-Himalayan region. Employing finite and hierarchical island population genetic models, we detected 30 (6.85%) loci under selection, of which 3.64% were under directional selection and 3.19% under balancing selection. The hierarchical island model was efficient in detecting signatures of natural selection with low occurrence of false positives. Data showed significant spatial genetic structure (SGS), especially in population from Nubra valley with significantly high SGS. Bayesian approach revealed high genetic differentiation and inbreeding in Nubra population. Indus and Suru valleys populations showed comparatively low SGS, genetic differentiation, and inbreeding. The study showed that M. alba falls in the range of mixed mating and outcrossing species. However, the population of Nubra valley showed selfing with high SGS. Significant SGS strengthens the presence of natural selection in M. alba natural populations.