Strategies aiming greater efficiency in the tobacco breeding program

Abstract

Tobacco crops have great socioeconomic importance in Brazil, where more than 90% of the Brazilian production is exported and thousands of families are involved with its cultivation. In this context, it is important to obtain information to improve the efficiency of the breeding programs. In tobacco recurrent selection (RS) different strategies can be employed in selection and recombination. Moreover, choosing the best strategy depends on the genetic variance components. Additionally, aiming to improve the efficiency of the breeding programs, the tobacco harvest is the process that demands the most labor and is also the process that has the highest possibility of contributing to experimental error. However, one of the alternatives that have been used in other species is phenotyping through images of the crop. Thus, the objective of this study was to estimate the genetic variance components in a tobacco population and compare the different strategies to conduct RS. Additionally, to verify the viability of employing the phenotyping of leaf mass (yield) through images collected by unmanned aerial vehicles (UAVs) and its likely applications in the tobacco breeding program. In the first study, it was found that the dominance variance was null for all evaluated traits, indicating that the dominance effect (d) is not important for the FCV tobacco population used in this study. In this condition, the estimate of 𝐷1, which depends on d, is not significant, meaning that inbreeding does not influence the gain from selection (𝐺𝑆) in the RS program; therefore, any type of progeny may be used. From the 𝐺𝑆 equation, it was found that the use of doubled haploid (DH) lines in tobacco RS represents the best strategy, as it associates greater 𝐺𝑆 per time unit and generates the line at the end of each cycle, which is the ultimate goal of the genetic breeding program. However, in the case of not using DH lines, other strategies allow expressive 𝐺𝑆, such as the alternate use of inbred and FS progenies for evaluation and FS progenies for recombination. In the second study, it was found that: i) image phenotyping presented good accuracy in detecting differences among the tobacco hybrids/lines, with the accuracy rising as the number of days after planting increased; ii) the canopy cover (CC), assessed through image phenotyping, correlated with the yield of green/cured leaves; iii) the employment of image phenotyping must be stimulated in tobacco crops aiming the obtainment of the green/cured mass yield; iv) it was conjectured that the application of images is viable in several situations within a breeding program aiming at the green/cured mass yield.