Genetic Progress in 53 Years of the Peach Breeding Program of Embrapa: Canning Genotypes


 Peach is a traditional crop in the south of Rio Grande do Sul State, Brazil, where 30 to 53 million cans of peaches in syrup are produced annually. All the raw material produced in the region consists of fruits originating from the peach breeding program of the Brazilian Agricultural Research Corporation (Embrapa Temperate Agriculture), which started even before Embrapa at the Experimental Station of Pelotas, Ministry of Agriculture. The objective was to estimate the genetic progress in phenological traits and production of canning peach resulting from the peach breeding program of Embrapa Temperate Agriculture in 53 years. We divided the data records considered in the estimation of genetic progress into two periods, 1964-1984 and 1985-2017, totaling 53 years. The following traits: maturing period, cycle, number of fruits, fruit weight, yield, and soluble solids content were evaluated. We initially tabulated data and analyzed descriptive statistics. Subsequently, we conducted analysis of mixed models and obtained the estimates of genetic progress through meta-analysis. Genetic gain for earliness, shortening the cycle from flowering to maturation, and genetic gain for fruit yield were observed.


66
The peach breeding program officially started in the municipality of Pelotas at the 4 other traits, not being included in the object of this study and not having the perspective of 101 becoming cultivars, we subsequently decided to only use cultivars that received variety 102 denomination. Nevertheless, the number of data was very large and we decided to separately 103 study table genotypes from those intended for canning (fruit produced for processing). 104 In the present study on genetic progress, we only considered canning peach genotypes 105 released by the Embrapa Temperate Agriculture program. Thus, we analyzed 52 canning  analysis. The two locations also differ in altitude (150 m and 60 m), which is reflected in the 116 temperature, generally 2 ºC higher at the ETC, consequently decreasing chilling accumulation 117 by 100 hours (Embrapa, 2020). 118 We analyzed genetic progress considering the genotypes listed in Table 1. 119 Data collection 120 The phenological monitoring of the cultivars is conducted yearly observing the dates of 121 beginning of flowering and full flowering and the beginning and end of maturation. Annually, 122 fruits from each cultivar are evaluated regarding fruit color, shape, and size, pulp firmness, total 123 soluble solids content, flavor, adherence to the endosperm, firmness, and health. 124 In the present study, we analyzed the following variables: beginning of maturation  136 We highlight that in the initial period, from 1964 to 1984, the number of fruits was 137 counted during fruit thinning. The number of fruits removed from the plant was counted and 138 the remaining fruits were counted afterwards. For this purpose, a manual counter was frequently 139 used to help record the numbers. However, fruit counting is very difficult in this period as the 140 plant presents many leaves and the fruits are green. Thus, failures may have occurred. In 141 addition, fruit may fall until harvest, which may have overestimated production at times in this 142 period. However, with the increase in the number of genotypes and reduction of human 143 resources to conduct this task, those counting became almost impossible in the second period.

144
Our option was to assign a degree of production on a scale of 1 to 5, where G1 = 50 fruits; G2 145 = 100 fruits; G3 = 250 fruits; G4 = 300 fruits; and G5 equal to or above 450 fruits. This scale 146 was attributed to be compatible with the 1964-1984 period, but it limited the largest 147 productions.

148
Data were recorded in a physical file. For analysis, they were digitized in an Excel®  where yij is the observed data of the variable in genotype i in year j, m is the intercept or 159 the general mean, gi is the random effect of genotype gi ~IIDN(0, 2 ), yj is the fixed effect of 160 year j and εij is the effect of the random error ε ~ IID N(0, 2 ).

161
The methodology for estimating genetic progress adopted was according to Breseghello   The variance components of random factors obtained by the generalized linear mixed 201 models are shown in Table 2. For the variance components of traits related to the cycle, we 202 observed a greater proportion of genotypic effects compared to environment effects. On the 203 other hand, environment variance strongly influenced production variables, that is, environment 204 effects provided great variations in the expression of these traits during the evaluated period.

205
For the fixed year effect, we observed significant effect on the factor for all the traits evaluated 206 in all periods according to the F test (Table 2).    and we highlight that the values were lower than for the first period ( Figure 2 a and b).    Table 4. 258 We observed that the slope was positive for maturing in the 1964-1984 period, resulting   Regarding the initial cycle (Figure 3 b), the genotypes `BRS Libra´ and `Pepita´  (Table 1). Thus, it is proved that in order to obtain the first very early-maturing

378
Regarding weight, the price differences between fruit considered of first and second 379 type class in the canneries, encourages the search for production of larger fruits, since yield and 380 profit is higher for the producer and yield in the production line is advantageous for the industry.

381
However, the so-called first quality fruit means being over 5.7 cm in diameter, and any 382 measurement above that has the same value. Diameter is highly correlated with fruit weight, 383 which means that there is no advantage in producing larger fruits beyond a certain average We observed genetic gain for earliness, with the harvest period extended to more than 418 three months.

419
There was anticipation of the beginning of the harvest and shortening in the cycle from 420 flowering to maturation, which did not affect fruit size as new cultivars became available from