Agronomic potential of BC₁F₂ populations of Santa Cruz dwarf tomato plants

ABSTRACT.

The use of dwarf lines to obtain mini-tomato hybrid plants has led to agronomic and economic advantages. However, the benefits provided by dwarf parents in tomato hybrids of the Santa Cruz type remain unexplored. The aims of this study were to determine the agronomic enhancement in BC₁F₂ dwarf populations bearing characteristic fruit of the Santa Cruz type after the first backcross and to select populations with high agronomic potential and fruit quality. The experiment was conducted in a greenhouse using a randomized block design with 15 treatments and 4 replicates. Evaluated genetic materials included 11 BC₁F₂ dwarf tomato populations, both parents (recurrent and donor), and the cultivars (Santa Cruz Kada and Santa Clara). Traits evaluated included: mean weight, soluble solid concentration, number of locules, shape, pulp thickness, fruit longitudinal and transversal diameters, internode length, and plant height. Univariate, multivariate, correlation, and selection index analyses showed that mean fruit weight, transverse diameter, and pulp thickness increased significantly in dwarf populations after a single backcross, most of them exhibiting a fruit shape similar to that of the Santa Cruz type. The dwarf BC₁F₂ populations UFU-Sci#11 and UFU-Sci#12 showed high potential for obtaining lines and, subsequently, Santa Cruz type hybrids.

Keywords:
Solanum lycopersicum L.; backcrossing; dwarf phenotype.

Introduction

Tomato (Solanum lycopersicum L.) is the second most grown vegetable crop worldwide (FAO, 2016). In 2020, approximately 55000 hectares of land was dedicated to tomato production in Brazil, with an estimated yield of 4 million tons (IBGE, 2020). In Brazil, tomatoes are classified into five categories: Minitomate, Salada, Caqui, Santa Cruz, and Saladete (Alvarenga, 2013). Outside Brazil, these tomato types are known as cherry or grape (mini-tomatoes), round, beefsteak, chonto, and saladette or roma, respectively (Finzi et al., 2020). Among them, the Santa Cruz fruit stands out with greater post-harvest durability, the highest yield potential, greater fruit uniformity, and a more intense red color (Shirahige, Melo, Purquerio, Carvalho, & Melo, 2010).

The cost of tomato production in Brazil is over 100,000 reals per hectare, and tomato cropping is considered a financially high-risk agricultural activity (Finzi et al., 2020). Future challenges in the sector include obtaining genotypes that are high yielding, resistant to pests and diseases, and show high fruit quality, thus requiring investments in innovation (Luz, Bittar, Oliveira, Nascimento, & Nogueira, 2016). The use of hybrid tomato plants has been the main strategy to achieve these objectives (Finzi, Maciel, Silva, Luz, & Borba, 2017a).

The use of dwarf plants is important for the development of tomato hybrids. Hybrid tomato plants obtained from a dwarf parent (obtained through crossing a normal plant with a dwarf plant) develop notably short internodes, thereby achieving a high yield by increasing the number of clusters per linear meter of stem (Finzi et al., 2017a). One cluster more per stem leads to a yield gain. As tomato is a crop that requires a high investment, adopting technologies aimed at increasing profit per hectare is fundamental to successful farming.

The dwarf line used by Finzi et al. (2017a) bears fruit that is incompatible for directly obtaining hybrids of the Santa Cruz tomato type, because it is small and oblong (Maciel, Silva, & Fernandes, 2015b). Finzi et al. (2020) performed the first backcross with a dwarf line mentioned aiming to obtain Salada type fruit. Similarly, backcrosses must be performed to obtain Santa Cruz type dwarf plant lines, which can then be used to obtain hybrids.

The aims of this study were to determine the agronomic enhancement in BC₁F₂ dwarf populations that bear fruit characteristic of the Santa Cruz type after the first backcross and to select populations with high agronomic potential and fruit quality. Thus, emphasis was given to comparisons between the dwarf populations and the donor parent.

Material and methods

The experiment was performed from March to September 2019 at the Hortaliças Experimental Station, Monte Carmelo, Minas Gerais State, Brazil (47º29’55.8” W; 18º42’43.19” S, 873 m asl). The plants were grown in an arch-type greenhouse (7 × 21 m) with a roof height of 4 m, covered with transparent, polyethylene film (150 microns) with additives to protect against ultraviolet radiation and with anti-aphid white screen lateral curtains.

The evaluated genetic material consisted of 15 treatments, including 11 populations of dwarf tomato plants obtained in the first backcross (BC₁F₂) after hybridization of a pre-commercial homozygous line bearing fruit of the Santa Cruz type (recurrent parent UFU-TOM-MOTHER-2) with the dwarf line (donor parent UFU MC TOM1; Maciel et al., 2015b), both parents, and the cultivars Santa Cruz Kada and Santa Clara.

The BC₁F₂ populations and the parents belong to the tomato germplasm bank of UFU. The recurrent parent and the cultivars (Santa Cruz Kada and Santa Clara) are characterized by an indeterminate growth habit and red Santa Cruz fruit type. The homozygous line UFU MC TOM1, used as the donor parent, includes dwarf plants with an indeterminate growth habit and oblong mini-tomato (cherry tomato) fruit type (Finzi, Maciel, Luz, Clemente, & Siquieroli, 2017b; Maciel et al., 2015b). Since expression of the dwarf phenotype is of recessive and monogenic origin (Maciel et al., 2015b), the backcrosses were performed for transferring the recessive allele.

Seeds were sown in 200 cell polystyrene trays on March 2^nd, 2019, and the seedlings were transplanted to 5 L plastic pots after 31 days. Seedlings of the respective dwarf tomato plant populations were selected within the BC₁F₂ populations before transplanting, as they were identified early because of their reduced structure and smaller, denser leaves, with dark green color. A coconut fiber-based commercial substrate was used in the trays and pots. Crop treatments were performed as recommended for the tomato crop grown in a protected environment (Alvarenga, 2013). The recurrent parent and the cultivars were oriented vertically with two stems in the polypropylene-cord-training system.

A randomized block experimental design was used, with 15 treatments and 4 replicates. Each experimental plot consisted of six plants distributed in double rows with a 0.3 × 0.3 m spacing. A spacing of 0.8 m was used between the double rows for a total of 360 plants.

Tomatoes were harvested weekly (a total of nine harvest time points) from June 5^th to August 2^nd, 2019. The fruit from each experimental plot was harvested at full maturity stage, and the following agronomic traits were evaluated:

Mean fruit weight (MFW) (g): the ratio of the sum total of individual fruit weights to the total number of fruit harvested from the plot.

Total soluble solid concentration (SSC) (ºBrix): the mean of all the fruit harvested in the plot, obtained using a portable digital refractometer (Atago PAL-1 3810).

Fruit transverse diameter (TD) (cm): the horizontal length measured with a ruler after cutting the fruit vertically in the middle.

Fruit longitudinal diameter (LD) (cm): the vertical length measured with a ruler after cutting the fruit vertically in the middle.

Fruit shape (FS): the ratio of the transverse diameter to the longitudinal diameter (TD/LD). The recurrent parent and the cultivars were used as references for the Santa Cruz segment to allow classification of the fruit.

Pulp thickness (PT) (cm): the length between the fruit peel and the beginning of the locule, measured with a ruler after cutting the fruit vertically in the middle.

Number of locules (NL) (locules fruit^-1): the number of locules counted after cutting the fruit horizontally in the middle.

Internode length (IL) (cm): obtained by the equation [(plant height/number of nodes)] based on the mean values of the two center plants of the plot.

Plant height (PH) (cm): determined by the vertical length of the plant as measured with a ruler and obtaining the mean value of the two center plants of the plot.

After verifying normality assumptions (Kolmogorov-Smirnov Test), homogeneity (O'Neil and Mathews Test), and additivity (Tukey’s Non-additivity Test), data transformation √(x+0.5) was used for mean fruit weight, number of locules, internode length, and plant height, registering the real values of these variables. In addition, the data obtained were analyzed with four independent models: means test, multivariate analyses, selection index, and correlations among traits. In the means test, analysis of variance was used (F = 0.05), and the mean values were compared using Tukey’s test (p = 0.05).

Multivariate analyses were performed solely for the purpose of determining the genetic dissimilarity between the dwarf populations and the donor parent; the dissimilarity matrix was obtained by measuring the generalized Mahalanobis distance. Genetic dissimilarity was represented by analyzing the canonical variables shown by scatter plot, with the axes represented by the first canonical variables (Cruz, Regazzi, & Carneiro, 2012).

The selection index was calculated by the rank-sum index of Mulamba and Mock (1978), using only the dwarf populations and the donor parent. Five populations were chosen for estimates of gains from selection. The selection criterion adopted was used to reduce internode length while increasing all other variables. The economic weight adopted was the genetic coefficient of variation of each variable, as recommended by Cruz et al. (2012). Analyses were performed using the Genes software (Cruz, 2016). The qgraph (Epskamp, Cramer, Waldorn, Schmittmann, & Borsboom, 2012) was used to obtain correlations that were processed by the R software version 3.6.3 (R Core Team, 2020).

Results and discussion

The BC₁F₂ dwarf tomato populations differed from the parents and the cultivars for all evaluated traits (Table 1), as per the F test (α = 0.05). Such traits exhibited heritability greater than 70% and a CVg/CVe ratio higher than 0.90 (Table 1), indicating the reliability of the selection process.

Table 1
Agronomic traits evaluated in BC₁F₂ populations of dwarf plants, recurrent parent, donor parent, and cultivars (Santa Cruz Kada and Santa Clara).

Treatments	MFW*	SSC	TD	LD	FS	PT	NL	IL	PH
UFU-Sci#2	7.25 de	9.66 ab	2.24 cd	3.04 de	0.74 cde	0.32 b	2.08 bc	1.30 ab	26.30 b
UFU-Sci#3	6.47 de	9.03 abc	2.30 cd	2.45 e	0.94 b	0.33 b	2.13 bc	1.30 ab	22.65 ab
UFU-Sci#5	5.98 de	8.39 bcd	2.43 c	2.70 e	0.90 bc	0.33 b	2.23 bc	1.22 ab	23.35 ab
UFU-Sci#6	4.46 e	8.95 abc	2.24 cd	2.39 e	0.93 b	0.31 b	2.29 bc	1.21 ab	24.45 ab
UFU-Sci#8	11.35 bc	8.96 abc	2.48 c	3.53 cd	0.70 de	0.35 b	2.35 bc	1.44 b	23.82 ab
UFU-Sci#11	8.29 cd	9.60 ab	2.35 cd	2.77 e	0.85 bcd	0.32 b	2.17 bc	1.25 ab	23.08 ab
UFU-Sci#12	8.34 cd	10.26 ab	2.20 cd	2.70 e	0.82 bcd	0.32 b	2.22 bc	1.11 ab	19.13 ab
UFU-Sci#16	7.22 de	9.86 ab	2.28 cd	2.60 e	0.88 bcd	0.30 b	2.10 bc	1.23 ab	23.88 ab
UFU-Sci#18	6.14 de	9.62 ab	2.17 cd	2.41 e	0.91 bc	0.32 b	2.03 c	1.27 ab	21.45 ab
UFU-Sci#20	6.19 de	9.59 ab	2.03 d	2.36 e	0.86 bcd	0.30 b	2.00 c	1.30 ab	21.03 ab
UFU-Sci#25	7.71 cd	8.97 abc	2.37 cd	2.71 e	0.87 bcd	0.32 b	2.11 bc	1.25 ab	26.58 b
Recurrent parent	13.90 ab	6.94 de	4.40 ab	4.82 a	0.91 bc	0.45 a	3.14 a	5.41 c	135.25 c
Donor parent	1.75 f	10.87 a	1.48 e	2.44 e	0.60 e	0.20 c	2.00 c	0.89 a	16.68 a
Kada	13.95 ab	7.32 cde	4.06 b	4.54 ab	0.89 bc	0.49 a	2.54 bc	5.65 c	185.00 d
Santa Clara	17.75 a	6.27 e	4.62 a	4.02 bc	1.16 a	0.45 a	2.55 ab	5.72 c	185.00 d
KS2	4.40	2.80	2.30	5.10	0.30	1.20	1.70	0.70	9.10
F (OM)3	0.90	1.90	0.30	8.60	1.80	0.40	1.50	0.90	1.40
F (Additivity)4	1.30	4.60	0.20	0.40	2.50	7.10	0.10	1.40	8.60
CV (%)	8.25	8.42	5.85	8.95	8.32	7.28	4.62	5.30	5.68

• *MFW: mean fruit weight (g), SSC: soluble solid concentration (ºBrix), TD: transverse diameter (cm); LD: longitudinal diameter (cm); FS: fruit shape; PT: pulp thickness (cm); NL: number of locules (locules fruit^-1); IL: internode length (cm); PH: plant height (cm). ¹Mean values followed by different letters within columns differ as per the Tukey test at 0.05 significance. KS², F (OM)³, F (Tukey)⁴: statistics of the Kolmogorov, O’Neil and Mathews, and Tukey tests, respectively. For all the variables evaluated, the residues exhibited normal distribution, homogeneous variances, and additivity between blocks and treatments.

Mean fruit weight and total number of tomatoes harvested are primary production components and have equal importance in determining total fruit production (Rodrigues, Marim, Silva, Mattedi, & Almeida, 2010). In the present study, fruit of the BC₁F₂ populations exhibited a considerable (312.6%) increase in mean weight, especially for the UFU-Sci#8 population (548.6%). This result shows the success of the first generation of backcross and the restoration of part of the genetic constitution of the recurrent parent. Similarly, Maciel et al. (2017) reported positive results after the first backcross in mini-tomato (cherry tomato).

Regarding soluble solid concentration, except for population UFU-Sci#5 (which scored a mean of 8.39°Brix), all other BC₁F₂ populations were similar to the donor parent, with a mean value of 9.45°Brix. Environmental, physiological, and genetic factors can modify the concentration of soluble solids. In this study, all evaluated genotypes were managed equally under greenhouse conditions, thus minimizing any environmental effect. Therefore, the low soluble solid concentration in UFU-Sci#5 is most likely because of genetic factors affecting fruit capacity to import photoassimilates. Finzi et al. (2019) evaluated the genetic variability for soluble solid concentration in mini-tomato hybrids derived from dwarf lines and mainly attributed the variation for the trait to genetic factors, consistently with our results.

For in natura consumption, soluble solid concentrations of 3.0°Brix is considered ideal (Kader, Morris, Stevens, & Albright-Holton, 1978; Schwarz et al., 2013). Thus, the BC₁F₂ populations are promising for the development of dwarf lines belonging to the Santa Cruz type, with high soluble solid concentration, and they can be used in enhancing current tomato plant varieties because the greater the soluble solid concentration, the greater the fruit sweet flavor, which is preferred by consumers (Maciel, Fernandes, Hillebrand, & Azevedo, 2015a; Schwarz et al., 2013).

Fruit size is represented in this study by transverse and lateral fruit diameters, which are important parameters related to the uniformity of the final product (Siddiqui, Ayala-Zavala, & Dhua, 2015). The mean increase in the transverse diameter of the fruit in the BC₁F₂ population (54.0%) suggests restoration of part of the genetic constitution of the recurrent parent, as observed for mean fruit weight. However, in the case of longitudinal diameter, except for the UFU-Sci#8 genotype (i.e., 3.53 cm), the BC₁F₂ populations did not differ from the donor parent, with mean values of 2.60 cm.

One of the factors for variability within the traits evaluated is the genotypic and phenotypic segregation after selection and subsequent self-fertilization of different BC₁F₁ plants (normal phenotype). To obtain the BC₁F₂ populations, BC₁F₁ plants were selected, followed by collection of seeds from BC₁F₂ populations. Thus, self-fertilization of the BC₁F₁ plants changed the genetic composition and reduced the number of heterozygous loci and, consequently, the expression of the traits. Consistently with the results reported herein, García-Fortea et al. (2019) reported segregation for the traits evaluated in the first two backcross generations in an eggplant crop, except for the number of flower parts.

Fruit transverse and lateral diameters also indicate an important trait of the tomato fruit shape, namely, the TD/LD ratio. All BC₁F₂ populations were similar to the recurrent parent and Santa Cruz Kada in this respect, exhibiting a slightly oblong shape. Therefore, clearly, the fruit borne by the evaluated BC₁F₂ populations were mostly similar to those of the Santa Cruz group. Fruit shape and size are related to firmness; consequently, when a fruit with an oval shape has increased pericarp thickness, it holds firm during post-harvest for a long period (Chakrabort, Vanlalliani, Chattopadhyay, & Hazra, 2007).

Pulp thickness in all BC₁F₂ populations was greater than that in the donor parent, showing an increase ranging from 55.0% (UFU-Sci#16 and UFU-Sci#20) to 75.0% (UFU-Sci#8). This trait is of fundamental importance to fruit quality, as a thicker pulp generally results in fruit with longer shelf life and better survival rate after long-distance transport (Chakrabort et al., 2007; Siddiqui et al., 2015). In this study, the BC₁F₂ populations exhibited a mean increase of 60.0% for PT in relation to the donor parent. In this step of the breeding program, the genotypes derived from the first backcross generation (BC₁F₂) are expected to exhibit a mean of 75.0% of the recurrent parent genome (Borém, Miranda, & Fritsche-Neto, 2017). Thus, this percentage is an indication of the success of the first backcross. However, other backcrosses are still necessary to increase the percentage of the genome of the recurrent parent in future lines.

Another factor affecting fruit quality is the number of locules; specifically, a higher number of locules is negatively associated with fruit firmness. Therefore, tomatoes with a small number of locules are desirable (Siddiqui et al., 2015). The number of locules did not differ among the individuals of the BC₁F₂ populations and the donor parent, which expressed a mean of 2.14 locules per fruit, thus, exhibiting a reduction in locules.

In addition to desirable fruit production and quality traits, obtaining compact plants, i.e., those with short internodes, should be one of the aims of tomato breeding programs. With a mean internode length of 1.20 cm and a mean plant height of 22.00 cm, all BC₁F₂ populations were similar to the donor parent, except for UFU-Sci#8 (1.44 cm) for internode length and UFU-Sci#2 and UFU-Sci#25, at 26.30 and 26.58 cm, for plant height, respectively. Phenotypic and genotypic segregation after selection and later self-fertilization of BC₁F₁ plants to obtain BC₁F₂ populations explained the variability observed among the different populations.

The reduced internode length of the dwarf populations is advantageous because it allows compact plants to be obtained in hybrid combinations, together with a large number of clusters per linear meter of stem (Finzi et al., 2017a). This trait facilitates pruning and vine training practices, thereby, reducing labor expenses and inputs (Marim et al., 2009; Figueiredo et al., 2015).

In general, one backcross was effective in increasing agronomic performance of the BC₁F₂ dwarf populations (Figure 1).

Figure 1
Comparison between the phenotypes of the donor and recurrent parents and the UFU-Sci#11 population with respect to agronomic parameters.

Other satisfactory results derived from the first backcross were observed in passion fruit (Melo, Souza, Sousa, Viana, & Santos, 2015), eggplant (García-Fortea et al., 2019), and tomato (Finzi et al., 2020). Thus, as in a previous study performed by Gonçalves Neto et al. (2010), performing two more backcrosses is suggested, as this will result in lines with commercial standard fruit and subsequent attainment of hybrids of the Santa Cruz type from dwarf lines, as obtained by Finzi et al. (2017a). This finding underlines the importance of using selection strategies as measurements of genetic dissimilarity, selection indices, and correlation among traits that allow superior dwarf lines to be obtained.

The dissimilarity estimated by the generalized Mahalanobis distance between dwarf plants ranged from 1.78 (UFU-Sci#11 and UFU-Sci#16) to 125.99 (donor parent and UFU-Sci#20), indicating genetic diversity among dwarf plants (data not shown). Visualization of the genetic dissimilarity was attained by a scatter plot of the canonical variables (Figure 2).

Figure 2
Scatter plot of the scores in relation to the two axes representing the first two canonical variables (VC1 and VC2). VC1 (95.23%) and VC2 (1.84%). 1 = UFU-Sci # 2; 2 = UFU-Sci # 3; 3 = UFU-Sci # 5; 4 = UFU-Sci # 6; 5 = UFU-Sci # 8; 6 = UFU-Sci # 11; 7 = UFU-Sci # 12; 8 = UFU-Sci # 16; 9 = UFU-Sci # 18; 10 = UFU-Sci # 20; 11 = UFU-Sci # 25; 12 = UFU-TOM-MOTHER-2 (recurrent parent); 13 = UFU MC TOM1 (donor parent); 14 = Santa Cruz Kada; and 15 = Santa Clara.

Using a scatter plot of the canonical variables, it was possible to identify three distinct groups (Figure 2). Group I was composed of the donor parent, Group II comprised all BC₁F₂ populations, and Group III consisted of the recurrent parent and the cultivars (Santa Cruz Kada and Santa Clara). The choice of the 2D graph was supported by the high percent values obtained from the first two canonical variables, which explained a great proportion of the variation within and among groups. For a satisfactory interpretation of the variability, it is necessary that the first two principal components explain over 80.0% of all the variability observed for the set of traits under study (Cruz et al., 2012). Here, the first two principal components explained 97.1% of the total variation.

The fact that all BC₁F₂ populations differed from the donor parent and had to be placed in different groups indicates the success of the first backcross. The two-dimensional plots have been widely used in different studies and explain the genetic variability among groups to a large extent (Gonçalves, Rodrigues, Amaral Junior, Karasawa, & Sudré, 2009; Oliveira, Silva, Brasileiro, Medeiros, & dos Anjos, 2013).

Normally, agronomic traits are subject to complex genetic control schemes affected by the environment, and strongly correlate with each other (Costa et al., 2004; Nick et al., 2013). Thus, selection considering only one trait is not very efficient when evaluating traits of polygenic inheritance and may result in gain for a certain trait but loss in other traits of agronomic importance (Cruz et al., 2012). According to Rezende et al. (2014) and Vasconcelos et al. (2010), selection of superior genotypes should be performed simultaneously for several traits, aiming at simultaneous genetic gain for the greatest possible number of traits. In this respect, the selection index is an excellent alternative, as it allows for simultaneous selection in an efficient manner (Cruz et al., 2012; Rosado, Santos, Bruckner, Nunes, & Cruz, 2012).

By the rank-sum index (Table 2), satisfactory gains were obtained for mean fruit weight (46.8%), soluble solid concentration (51.2%), fruit shape (10.5%), and number of locules (27.4%). Moreover, we estimated that there was a reduction in internode length (-30.4%), a trait favorable for tomato (Finzi et al., 2017a), as an additional aim was to obtain more compact plants. Finzi et al. (2020) evaluated selection of tomato lines using the Mulamba and Mock selection index and, consistently with our results, reported a reduction in internode length (-2.39%). The greatest genetic gain, evenly distributed among all the evaluated traits, were obtained through selection of five populations: UFU-Sci#3, UFU-Sci#11, UFU-Sci#12, UFU-Sci#20, and UFU MC TOM1 (Table 2).

Table 2
Estimate gain from selection (GS%) obtained for nine traits among 12 dwarf tomato populations (11 BC₁F₂ populations and the donor parent UFU MC TOM1) by the Mulamba and Mock rank-sum index.

Selected genotypes
Traits	%GS
MFW	46.8
SSC	51.2
TD	-26.5
LD	-10.9
FS	10.5
PT	-39.0
NL	27.4
IL	-30.4
PH	-58.0
%GS Total	-28.9
Selected Populations	UFU-Sci#3
	UFU-Sci#11
	UFU-Sci#12
	UFU-Sci#20
	UFU MC TOM1

• MFW: mean fruit weight (g), SSC: soluble solid concentration (ºBrix), TD: transverse diameter (cm); LD: longitudinal diameter (cm); FS: fruit shape; PT: pulp thickness (cm); NL: number of locules (locules fruit^-1); IL: internode length (cm); and PH: plant height (cm).

Additional important information was obtained from the relationship among traits, i.e., genotypic, phenotypic, and environmental correlations (Ferreira, Queiróz, Braz, & Vencovsky, 2003; Valadares et al., 2017). This estimate is emphasized especially when difficulties emerge in the selection for a desirable trait because of low heritability and/or difficult measurement and identification, thus, allowing for gains to be obtained for one trait through indirect selection for another (Cruz et al., 2012; Valadares et al., 2017). Positive correlation coefficients indicate the tendency for a variable to increase when a correlated variable increases, whereas negative values indicate the tendency for a variable to decrease when another increases (Nogueira et al., 2012). In the present study, strong, positive genetic correlations were observed between the pairs: (NL × LD), (TD × LD), (IL × LD), (IL × TD), (PH × IL), (MFW × LD), (MFW × TD), and (PT × TD) (Figure 3). However, all traits had a negative correlation coefficient when correlated with soluble solid concentration. According to Falconer and Mackay (1996), and Nogueira et al. (2012), pleiotropy (i.e., the same gene affects the expression of more than one trait), is one of the causes for high correlations. This information can contribute to advances in tomato breeding programs by increasing the efficiency of the selection process and facilitating the choice of parents for future crosses.

Figure 3
Genotypic correlations among nine traits in tomato. MFW: mean fruit weight (g), SSC: soluble solid concentration (ºBrix), TD: transverse diameter (cm); LD: longitudinal diameter (cm); FS: fruit shape; PT: pulp thickness (cm); NL: number of locules (locules fruit^-1); IL: internode length (cm); PH: plant height (cm). X: non-significant correlation. The area in blue represents positive correlation, whereas that in red represents negative correlation. The darker the color, the greater the correlation.

Backcross populations UFU-Sci#11 and UFU-Sci#12 were prominent in both, univariate analysis and in the estimate of genetic gain. In addition, the scatter plot for the canonical variables helped confirm genetic dissimilarity in relation to the donor parent. This indicates that UFU-Sci#11 and UFU-Sci#12 have a high agronomic potential. Thus, we propose using these populations for further backcrossing to obtain hybrids from dwarf lines of the Santa Cruz tomato type.

Conclusion

A single backcross allowed us to obtain dwarf plant populations with a significant increase in agronomic performance for the Santa Cruz tomato type. Positive and strong genetic correlations were observed between the traits (NL × LD), (TD × LD), (IL × LD), (IL × TD), (PH × IL), (MFW × LD), (MFW × TD), and (PT × TD). The populations UFU-Sci#11 and UFU-Sci#12 are recommended for obtaining lines and, subsequently, for hybrids of the Santa Cruz type of tomato.

Acknowledgements

The authors gratefully acknowledge financial support by Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) and Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

References

• Alvarenga, M. A. R. (2013). Tomate: produção em campo, em casa-de-vegetação e em hidroponia (2. ed.). Lavras, MG: Editora UFLA.
• Borém, A., Miranda, G. V., & Fritsche-Neto, R. (2017). Melhoramento de plantas (6. ed.). Viçosa, MG: Editora UFV.
• Chakraborty, I., Vanlalliani,V., Chattopadhyay, A., & Hazra, P. (2007). Studies on processing and nutritional qualities of tomato as influenced by genotypes and environment. Vegetable Science, 34(1), 26-31.
• Costa, M. M., Di Mauro, A. O., Unêda-Trevisoli, S. H., Arriel, N. H. C., Bárbaro, I. M., & Muniz, F. R. S. (2004). Ganho genético por diferentes critérios de seleção em populações segregantes de soja. Pesquisa Agropecuária Brasileira, 39(11), 1095-1102. DOI: https://doi.org/10.1590/S0100-204X2004001100007
  » https://doi.org/https://doi.org/10.1590/S0100-204X2004001100007
• Cruz, C. D., Regazzi, A. J., & Carneiro, P. C. S. (2012). Modelos biométricos aplicados ao melhoramento genético (4. ed.). Viçosa, MG: Editora UFV .
• Cruz, C. D. (2016). Genes Software - extended and integrated with the R, Matlab and Selegen. Acta Scientiarum Agronomy, 38(4), 547-552. DOI: https://doi.org/10.4025/actasciagron.v38i4.32629
  » https://doi.org/https://doi.org/10.4025/actasciagron.v38i4.32629
• Epskamp, S., Cramer, A. O., Waldorn, L. J., Schmittmann, V. D., & Borsboom, D. (2012). qgraph: Network visualizations of relationships in psychometric data. Journal of Statistical Software, 48(4), 1-18. DOI: https://doi.org/10.18637/jss.v048.i04
  » https://doi.org/https://doi.org/10.18637/jss.v048.i04
• Falconer, D. S., & Mackay, T. F. C. (1996). Introduction to quantitative genetics Edinburgh, SC: Addison Wesley Longman.
• Food and Agriculture Organization of the United Nations [FAO]. (2016). Production of food and agricultural products Retrieved on Aug. 1, 2020 from 1, 2020 from http://faostat.fao.org
  » http://faostat.fao.org
• Ferreira, M. A. J., Queiróz, M. A. D., Braz, L., & Vencovsky, R. (2003). Correlações genotípicas, fenotípicas e de ambiente entre dez caracteres de melancia e suas implicações para o melhoramento genético. Horticultura Brasileira, 21(3), 438-442. DOI: https://doi.org/10.1590/S0102-05362003000300004
  » https://doi.org/https://doi.org/10.1590/S0102-05362003000300004
• Figueiredo, A. S. T., Meert, L., Paula, J. T., Resende, J. T. V., Rodrigues, J. D., & Ono, E. O. (2015). Comportamento de plantas de tomateiro indeterminado na presença de regulador de crescimento. Revista Campo Digital, 10(1), 31-40.
• Finzi, R. R., Maciel, G. M., Silva, E. C. D., Luz, J. M. Q., & Borba, M. E. A. (2017a). Agronomic performance of mini-tomato hybrids from dwarf lines. Ciência e Agrotecnologia, 41(1), 15-21. DOI: https://doi.org/10.1590/1413-70542017411021416
  » https://doi.org/https://doi.org/10.1590/1413-70542017411021416
• Finzi, R. R., Maciel, G. M., Luz, J. M. Q., Clemente, A. A., & Siquieroli, A. C. S. (2017b). Growth habit in mini tomato hybrids from a dwarf line. Bioscience Journal, 33(1), 52-56. DOI: https://doi.org/10.14393/Bj-v33n1a2017-35763
  » https://doi.org/https://doi.org/10.14393/Bj-v33n1a2017-35763
• Finzi, R. R., Marquez, G. R., Maciel, G. M., Momesso, M. P., Pereira, L. M., & Silveira, A. J. (2019). Soluble solids due to the truss position in minitomato hybrids from dwarf lines. Agrarian, 12(43), 33-39. DOI: https://doi.org/10.30612/agrarian.v12i43.8948
  » https://doi.org/https://doi.org/10.30612/agrarian.v12i43.8948
• Finzi, R. R., Maciel, G. M., Peres, H. G., Silva, M. F., Peixoto, J. V. M., & Gomes, D. A. (2020). Agronomic potential of BC₁F₂ dwarf round tomato populations. Ciência e Agrotecnologia , 44(2), 1-8. DOI: https://doi.org/10.1590/1413-7054202044028819
  » https://doi.org/https://doi.org/10.1590/1413-7054202044028819
• García-Fortea, E., Gramazio, P., Vilanova, S., Fita, A., Mangino, G., Villanueva, G., ... Plazas, M. (2019). First successful backcrossing towards eggplant (Solanum melongena) of a New World species, the silverleaf nightshade (S. elaeagnifolium), and characterization of interspecific hybrids and backcrosses. Scientia Horticuturae, 246(4), 563-573. DOI: https://doi.org/10.1016/j.scienta.2018.11.018
  » https://doi.org/https://doi.org/10.1016/j.scienta.2018.11.018
• Gonçalves Neto, Á. C., Silva, V. D. F., Maluf, W. R., Maciel, G. M., Nízio, D. A., Gomes, L. A., & Azevedo, S. M. (2010). Resistência à traça-do-tomateiro em plantas com altos teores de acilaçúcares nas folhas. Horticultura Brasileira , 28(2), 203-208. DOI: https://doi.org/10.1590/S0102-05362010000200011
  » https://doi.org/https://doi.org/10.1590/S0102-05362010000200011
• Gonçalves, L. S. A., Rodrigues, R., Amaral Junior, A. T., Karasawa, M., & Sudré, C. P. (2009). Heirloom tomato gene bank: assessing genetic divergence based on morphological, agronomic and molecular data using a Ward-modified location model. Genetics and Molecular Research, 8(1), 364-374. DOI: https://doi.org/10.4238/vol8-1gmr549
  » https://doi.org/https://doi.org/10.4238/vol8-1gmr549
• Instituto Brasileiro de Geografia e Estatística [IBGE]. (2020). Levantamento sistemático da produção agrícola Retrieved on Jan. 18, 2020 from 18, 2020 from https://sidra.ibge.gov.br/home/lspa/brasil
  » https://sidra.ibge.gov.br/home/lspa/brasil
• Kader, A. A., Morris, L. L., Stevens, M. A., & Albright-Holton, M. (1978). Composition and flavor quality of fresh market tomatoes as influenced by some postharvest handling procedures. Journal of the American Society for Horticultural Science, 103(1), 6-13.
• Luz, J. M. Q., Bittar, C. A., Oliveira, R. C., Nascimento, A. R., & Nogueira, A. P. (2016). Desempenho e divergência genética de genótipos de tomate para processamento industrial. Horticultura Brasileira , 34(4), 483-490. DOI: https://doi.org/10.1590/s0102-053620160406
  » https://doi.org/https://doi.org/10.1590/s0102-053620160406
• Maciel, G. M., Almeida, R. S., Rocha, J. P., Andaló, V., Marquez, G. R., Santos, N. C., & Finzi, R. R. (2017). Mini tomato genotypes resistant to the silverleaf whitefly and to two-spotted spider mites. Genetics and Molecular Research , 16(1), 1-9. DOI: https://doi.org/10.4238/gmr16019539
  » https://doi.org/https://doi.org/10.4238/gmr16019539
• Maciel, G. M., Fernandes, M. A. R., Hillebrand, V., & Azevedo, B. N. R. (2015a). Influência da época de colheita no teor de sólidos solúveis em frutos de minitomate. Scientia Plena, 11(12), 1-6. DOI: https://doi.org/10.14808/sci.plena.2015.120203
  » https://doi.org/https://doi.org/10.14808/sci.plena.2015.120203
• Maciel, G. M., Silva, E. C. D., & Fernandes, M. A. R. (2015b). Ocorrência de nanismo em planta de tomateiro do tipo grape. Revista Caatinga, 28(4), 259-264. DOI: https://doi.org/10.1590/1983-21252015v28n429rc
  » https://doi.org/https://doi.org/10.1590/1983-21252015v28n429rc
• Marim, B. G., Silva, D. J. H. D., Carneiro, P. C. S., Miranda, G. V., Mattedi, A. P., & Caliman, F. R. B. (2009). Variabilidade genética e importância relativa de caracteres em acessos de germoplasma de tomateiro. Pesquisa Agropecuára Brasileira, 44 (10), 1283-1289. DOI: https://doi.org/10.1590/S0100-204X2009001000011
  » https://doi.org/https://doi.org/10.1590/S0100-204X2009001000011
• Melo, C. A. F., Souza, M. M., Sousa, A. G. R., Viana, A. P., & Santos, E. A. (2015). Multivariate analysis of backcross progeny of Passiflora L. (Passifloraceae) for pre-breeding genotype selection. Genetics and Molecular Research , 14(4),15376-15389. DOI: https://doi.org/10.4238/2015.November.30.15
  » https://doi.org/https://doi.org/10.4238/2015.November.30.15
• Mulamba, N. N., & Mock, J. J. (1978). Improvement of yield potential of the Eto Blanco maize (Zea mays L.) population by breeding for plant traits. Egyptian Journal of Genetics and Citology, 7(1), 40-41.
• Nick, C., Laurindo, B. S., Almeida, V. D. S., Freitas, R. D. D., Aguilera, J. G., Silva, E. C. F. D., ... Silva, D. J. H. D. (2013). Seleção simultânea para qualidade do fruto e resistência à requeima em progênies de tomateiro. Pesquisa Agropecuária Brasileira , 48(1), 59-65. DOI: https://doi.org/10.1590/S0100-204X2013000100008
  » https://doi.org/https://doi.org/10.1590/S0100-204X2013000100008
• Nogueira, A. P. O., Sediyama, T., Sousa, L. B., Hamawaki, O. T., Cruz, C. D., & Pereira, D. G. (2012). Análise de trilha e correlações entre caracteres em soja cultivada em duas épocas de semeadura. Bioscience Journal , 28(6), 877-888.
• Oliveira, R. S., Silva, S. A., Brasileiro, B. P., Medeiros, E. P., & dos Anjos, E. V. A. (2013). Divergência genética em mamoneira utilizando a estratégia ward-mlm. Revista Ciência Agronômica, 44 (3), 564-570. DOI: https://doi.org/10.1590/S1806-66902013000300019
  » https://doi.org/https://doi.org/10.1590/S1806-66902013000300019
• R Core Team. (2020). R: A language and environment for statistical computing Vienna, AT: R Foundation for Statistical Computing. Retrieved on Feb. 27, 2020 from 27, 2020 from http://www.R-project.org
  » http://www.R-project.org
• Rezende, J. C. D., Boltelho, C. E., Oliveira, A. C. B. D., Silva, F. L. D., Carvalho, G. R., & Pereira, A. A. (2014). Genetic progress in coffee progenies by different selection criteria. Coffee Science, 9(3), 347-353. DOI: https://doi.org/10.25186/cs.v9i3.643
  » https://doi.org/https://doi.org/10.25186/cs.v9i3.643
• Rodrigues, G. B., Marim, B. G., Silva, D. J. H. D., Mattedi, A. P., & Almeida, V. D. S. (2010). Análise de trilha de componentes de produção primários e secundários em tomateiro do grupo Salada. Pesquisa Agropecuária Brasileira , 45(2), 155-162. DOI: https://doi.org/10.1590/S0100-24X2010000200006
  » https://doi.org/https://doi.org/10.1590/S0100-24X2010000200006
• Rosado, L. D. S., Santos, C. E. M. D., Bruckner, C. H., Nunes, E. S., & Cruz, C. D. (2012). Simultaneous selection in progenies of yellow passion fruit using selection indices. Revista Ceres, 59(1), 95-101. DOI: https://doi.org/10.1590/S0034-737X2012000100014
  » https://doi.org/https://doi.org/10.1590/S0034-737X2012000100014
• Schwarz, K., Resende, J. T. V., Preczenhak, A. P., Paula, J. T., Faria, M. V., & Dias, D. M. (2013). Desempenho agronômico e qualidade físico-química de híbridos de tomateiro em cultivo rasteiro. Horticultura Brasileira , 31(3), 410-418. DOI: https://doi.org/10.1590/S0102-05362013000300011
  » https://doi.org/https://doi.org/10.1590/S0102-05362013000300011
• Shirahige, F. H., Melo, A. M. T., Purquerio, L. F. V., Carvalho, C. R. L., & Melo, P. C. T. (2010). Produtividade e qualidade de tomates Santa Cruz e Italiano em função do raleio de frutos. Horticultura Brasileira , 28(3), 292-298. DOI: https://doi.org/10.1590/S0102-05362010000300009
  » https://doi.org/https://doi.org/10.1590/S0102-05362010000300009
• Siddiqui, M. W., Ayala-Zavala, J. F., & Dhua, R. S. (2015). Genotypic variation in tomatoes affecting processing and antioxidant properties. Critical Reviews in Food Science and Nutrition, 55(13), 1819-1835. DOI: https://doi.org/10.1080/10408398.2012.710278
  » https://doi.org/https://doi.org/10.1080/10408398.2012.710278
• Valadares, R. N., Melo, R. A., Silva, J. A. S., Araújo, A. L. R., Silva, F. S., Carvalho Filho, J. L. S., & Menezes, D. (2017). Estimativas de parâmetros genéticos e correlações em acessos de melão do grupo momordica. Horticultura Brasileira , 35(4), 557-563. DOI: https://doi.org/10.1590/s0102-053620170413
  » https://doi.org/https://doi.org/10.1590/s0102-053620170413
• Vasconcelos, E. S., Ferreira, R. D. P., Cruz, C. D., Moreira, A., Rassani, J. B., & Freitas, A. R. D. (2010). Estimativas de ganho genético por diferentes critérios de seleção em genótipos de alfafa. Revista Ceres , 57(2), 205-210. DOI: https://doi.org/10.1590/S0034-737X2010000200011
  » https://doi.org/https://doi.org/10.1590/S0034-737X2010000200011

Publication Dates

• Publication in this collection
  28 Oct 2022
• Date of issue
  2023

History

• Received
  02 Nov 2020
• Accepted
  14 Apr 2021