Estudos e avanços na estratégia de seleção recorrente para maximizar o ganho genético em arroz

Abstract

Upland rice is mainly cultivated in the cerrado, where the soils have a low phosphorus content and a high aluminum content. The limited availability of phosphorus as a fertilizer, due to the abundant presence of iron and aluminum oxides, hinders nutrition and impacts agricultural yields. Faced with these challenges, the identification of genotypes more efficient in phosphorus use stands out as a low-cost strategy. However, increasing the yield potential in upland rice is faced with abiotic stresses represents one of the main challenges in breeding programs. Additionally, to achieve significant genetic progress is becoming increasingly challenging. Thus, one alternative to maintain and broaden genetic variability, to obtain new recombinations in breeding programs, and accelerate long-term genetic gains is through recurrent selection. Using the recurrent selection population CNA9 of upland rice, the present study aims: (i) From 3th cycle to 4th cycle: to estimate the genetic progress for grain yield (GY), plant height (PH), and days to flowering (DH) conducted under contrasting soil phosphorus levels. Since 5th cycle: (ii) To estimate the genetic and phenotypic parameters between two S 0:2 progenies and S 2 half-siblings of the population;; (iii) To estimate the expected response to selection and the annual genetic gain for grain yield (GY), plant height (PH) and days to flowering (DF), for each method (iv) To validate the new method of recurrent selection in rice using male genetic sterility. In each cycle, trials were carried out under low and high soil P levels in two locations: Santo Antônio de Goiás, GO, and Sinop, MT. The third cycle were performed with 189 progenies in simple square lattice 14 x 14, while the fourth cycle with 137 progenies in simple square lattice 12 x 12, both with the same seven cultivars as checks. In the 5 th cycle two experiments using S 0:2 and S2 half-sib progenies were carried out in Lavras/MG (2019/2020 and 2020/2021) and Santo Antônio de Goiás/GO (2020/2021). The S 0:2 population consisted of 210 genotypes in a simple alpha lattice 18x12 design. On the other hand, the S 2 hal-sib consisted of 42 progenies in a simple alpha lattice 8x6 design. With the measured values from 3th cycle to 4th, the results showed genetic progress for all traits, mainly GY in high and low soil P levels, with respective estimates of annual genetic gains of 10.31% and 2.78% between the third and fourth cycles of recurrent selection, indicating the effectiveness of the selection criteria for this trait. Furthermore, there was a favorable maintenance of the means for PH and DF between these two cycles. In the 5 th cycle, estimates of the genetic and phenotypic parameters showed genetic variability in both, the S 0:2 and S 2 half-sib progenies, for the traits DF, PH, and GY. Annual genetic gain (%) for GY and DF of S 2 half-sib was 2-fold higher than those of S 0:2 , whereas, for PH, the annual genetic gain % of S 2 half-sib progenies was 4-fold higher than S0:2. Considering the feasibility of reducing the duration of each three-year selection cycle to two years through transitioning from using S 0:2 to S 1 half-sib progenies, a positive impact on the increase in annual genetic gain (%) was observed, demonstrating the efficiency of the new strategy.