Analysis strategies for optimizing the selection of potato clones

Abstract

Potato (Solanum tuberosum L.) stands out as the fourth fundamental crop for global food security, behind rice, wheat, and corn, due to its high productive potential per area, short cycle, and nutritional value. However, genetic improvement of this crop faces significant challenges due to tetrasomic inheritance, high parental heterozygosity, low tuber multiplication rate, low heritability for tuber yield traits, small genetic gains, and genotype-by-environment interaction (G×E). With the objective of improving the efficiency of potato genetic improvement, two studies were conducted to promote strategies to increase the selective accuracy of potato clones. The first investigated the contribution of the family effect in the selection of clones in the initial phase of a potato breeding program involving clonal evaluation in Single Trial (ST) or Multi-Environment Trials (MET). 1280 clones from 67 full-sib families (biparental crosses) under different heat stress conditions were evaluated for the traits total tuber yield (TTY) and specific gravity (SG). The second analyzed the partition of yield components and their impact on selective accuracy, evaluating 620 clones using Single-trait Mixed Model (STMM) and Multi-trait Mixed Model (MTMM) in the winter crop (without heat stress) for the traits average tuber mass (ATM), total tuber number (TTN), and total tuber yield (TTY), using a digital imaging approach. The results demonstrated that the inclusion of the family effect increased selective efficiency in both ST and MET analyses. The MTMM model showed better fit than STMM, but showed no increase in selective accuracy in univariate analysis for the winter crop. However, multivariate analysis for potato yield components can be useful in scenarios with low environmental control, providing increased selective accuracy of clones in early stages of the potato breeding program. The inclusion of the family effect is advantageous, as it increases the selective accuracy of potato clones without incurring additional costs. Keywords: Solanum tuberosum L.; selective accuracy; multivariate analysis; computer vision.