Digital phenotyping for quantification of genetic diversity in inbred guava (Psidium guajava) families

Digital image analysis of seeds has been used for the identification of cultivars, determination of seed color and mechanical damage, and classification of different seed sizes. The aim of the present study was to evaluate the efficiency of digital image analysis of seeds for the quantification of genetic diversity among genotypes of inbred guava (Psidium guajava L.) families. The SAS Mini equipment, which consists of a capture module and a software program for analysis, was employed for the capture and analysis of the seed images. Different genetic diversity quantification strategies were tested using the Ward-Modified Location Model method. The set of variables related to geometry of the seeds was the largest contributor to divergence among the guava genotypes. The use of seed descriptors obtained by digital image analysis via the SAS system was efficient at quantifying the genetic diversity among genotypes of inbred guava families associated with the use of the Ward-Modified Location Model method.