Agromorphological Evaluation of Elite Lines of Native Tomato (Solanum lycopersicum L.) from Central and Southern Mexico

Abstract

Tomato (Solanum lycopersicum L.) is one of the most important cultivated vegetables in the world. However, in some countries such as Mexico the lack of cultivars adapted to different environmental production conditions is a limitation. Moreover, recent studies have indicated that breeding aimed at increasing yield has led to a loss of genetic diversity. Therefore, it is necessary to explore and characterize new sources of germplasms. This study aimed to characterize new sources of germplasm and identify the most transcendental traits for distinguishing tomato types and lines that are useful for the genetic improvement of the species. Sixty characters were evaluated in 16 advanced lines of native tomatoes from Central and Southern Mexico during the fall–winter cycles 2023–2024 at the Bajío Experimental Station, Celaya, Guanajuato, Mexico, based on the guidelines of the International Union for the Protection of New Varieties of Plants (UPOV) and the International Plant Genetic Resources Institute (IPGRI). The data were analyzed using descriptive statistics, analysis of variance and post hoc tests, canonical discriminant analysis, and the Eigenanalysis selection index method (ESIM). Morphological variation showed that five qualitative traits were determinant factors in distinguishing tomato types and lines, whereas agronomic discriminant traits were the equatorial and polar diameters of the fruit and its ratio, number of locules, pedicel length, stem length, and internode distance. In addition, significant positive correlations were found between leaf length and width, equatorial diameter of the fruit, and polar diameter of the fruit. Lines JCM-17, JMC-10, and JCM-01 were the most selectable lines according to the ESIM values. The morphological variation found and the characteristics with higher selection values identified may be valuable for optimizing the tomato genetic improvement process in general.

1. Introduction

Agriculture is an important global activity, not only for food production but also for generating employment, foreign exchange, and its relations with the production and economic chain [1]. Among economically important vegetables, tomato (Solanum lycopersicum L.) is one of the three most cultivated vegetables worldwide [2]. China produces 34% of the world’s tomato crop [3], whereas the USA is the main importer, with Mexico as its principal supplier, which focuses 99.7% of its production in this market [4]. However, tomato production in Mexico faces several challenges, including phytosanitary issues, nutritional plant imbalance, conservation difficulties, marketing problems, and, perhaps most importantly, a lack of improved cultivars adapted to the diverse agroecological conditions and production systems in which it is grown [5].

Despite yield improvements achieved through domestication, breeding efforts focused on increasing productivity have been accompanied by a loss of genetic diversity and a decrease in nutritional value and fruit flavor [6]. Several authors have hypothesized that for millennia, breeding efforts have focused on these specific traits, resulting in the reduction in genetic diversity and the loss of other beneficial traits possessed by wild relatives, such as tolerance to pests, diseases, and abiotic stresses [6,7]. The loss of diversity among cultivated species poses a risk to sustainable production, particularly when most cultivars share the same genetic base [7]. Therefore, it is crucial to take advantage of the traits that native varieties develop in response to adverse biotic and abiotic factors [8].

The loss of genetic variation in modern tomato cultivars has limited their breeding potential [9], which can be addressed using the available diversity, mainly in the important agronomic traits of native varieties and wild relatives [10]. In this context, the sensorial and other agronomic traits of native varieties can be valuable for developing hybrid cultivars [11].

Central and Southern Mexico are considered the center of diversity and domestication of tomatoes. Hence, there is a useful native germplasm to improve cultivars, which is particularly important given that tomato production in Mexico largely depends on imported seeds [12]. This dependency is mainly due to limited breeding progress stemming from a lack of information about the morphological, phenological, and biochemical traits of the native and wild germplasm of this species. Thus, it is a priority to research, explore, characterize, and exploit the beneficial traits of native germplasm to incorporate important agronomic traits into improved cultivars.

Currently, Mexican studies on the morphological characterization of tomatoes are limited and mainly focus on describing fruit traits such as shape, size, color intensity, and the presence of a green shoulder, among others [13,14,15]. Therefore, it is necessary to expand the knowledge on tomato morphological diversity in all plant organs in a larger number of accessions to contribute to tomato breeding programs in Mexico. The guidelines proposed by the International Union for the Protection of New Varieties of Plants (UPOV) and the International Plant Genetics Resources Institute (IPGRI) are the basis for characterizing these native germplasms. Additionally, multivariate statistical analyses and mathematical selection indices can provide a comprehensive approach to tomato breeding [16,17,18].

Given these considerations, this study aimed to characterize 16 elite lines of native tomatoes from Central and Southern Mexico to identify germplasm differences and the principal traits of variation. Furthermore, it sought to calculate the ESIM selection index using important agronomical traits to provide an objective approach for tomato selection.

2. Materials and Methods

2.1. Experimental Site, Plant Material, and Experimental Design

This research was carried out in a greenhouse of the vegetable program at the Bajío Experimental Station of the Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias (INIFAP), located 6.5 kilometers on the Celaya-San Miguel de Allende highway, Celaya, Guanajuato, Mexico (20°32′53.75″ N latitude, 100°49′13″ W longitude, and an altitude of 1752 m above sea level). This location has a semi-dry and semi-warm climate (BS_W) with summer rains, an annual rainfall of 600–700 mm, and an average temperature of 17.6 °C [19].

Sixteen elite lines (F5) of native tomatoes (Solanum lycopersicum L.) from the germplasm bank of the vegetable breeding program at Bajío Experimental Station were evaluated (

Table 1. Fruit shape and origin of the populations from which the sixteen elite lines were segregated.

Line ID	Origin	Fruit Shape	North Latitude	West Longitude	Altitude (m)
JCM-01	Dzitbalche, Campeche	Kidney	20°19′18.30″	90°03′29.59″	17
JCM-02	Huachinango, Puebla	Kidney	20°10′20.25″	98°03′44.66″	1.565
JCM-03	Tlacolula, Oaxaca	Kidney	16°57′04.32″	96°28′28.67″	1.631
JCM-04	Poza Rica, Veracruz	Kidney	20°31′18.81″	97°27′45.54″	67
JCM-05	Xoxocotla, Morelos	Cherry	18°40′52.69″	99°14′46.61″	1.023
JCM-06	Tlacolula, Oaxaca	Cherry	16°57′04.32″	96°28′28.67″	1.631
JCM-07	Teapa, Tabasco	Kidney	17°33′52.71″	92°57′00.39″	40
JCM-08	Dzitbalche, Campeche	Half kidney	20°19′18.30″	90°03′29.59″	17
JCM-09	Tlacolula, Oaxaca	Saladette	16°57′04.32″	96°28′28.67″	1.631
JCM-10	Huachinango, Puebla	Kidney	20°10′20.25″	98°03′44.66″	1.565
JCM-11	Zitlala, Puebla	Kidney	20°01′57.91″	97°40′04.53″	732
JCM-12	Zozocolco de Hidalgo, Veracruz	Kidney	20°08′22.30″	97°34′30.02″	305
JCM-15	Altepexi, Puebla	Creole chino	18°21′27.55″	97°18′01.51″	1.229
JCM-16	San Sebastián Zinacatepec, Puebla	Creole chino	18°20′04.21″	97°14′46.67″	1.143
JCM-17	San José Miahuatlán, Puebla	Creole chino	18°17′25.50″	97°17′22.36″	1.122
JCM-18	San Sebastián, Tehuacán, Puebla	Creole chino	18°27′39.89″	97°24′25.47″	1.658

).

Each elite line was initially segregated using the mass selection method based on fruit shape, resulting in five fruit shape types: kidney, half-kidney, cherry, saladette, and creole chino. Subsequently, the sixteen elite lines were segregated by individual selection, yielding eight kidney lines, two cherry lines, one half-kidney line, one saladette line, and four creole chino lines.

A Randomized Complete Block (RCB) design with 16 lines (treatments) and 4 replicates was used to control for potential nuisance factors that may arise in the greenhouse. Each experimental unit had 10 plants distributed in a 4.0 m² zigzag topological arrangement in a double row, with 1.0 m between rows and 0.5 m between plants. Seedling production, transplantation, and crop management were performed according to technological recommendations for tomato production by INIFAP researchers [20]. Pollination was performed mechanically by shaking the plants twice daily (8:00 and 12:00 h) using the raffia of the tutoring system.

2.2. Data Collected

Sixty traits based on the UPOV [21] and IPGRI [22] guidelines were evaluated, including the following: (1) seedling: anthocyanin coloration of hypocotyl; (2) plant: growth type; (3) stem: anthocyanin coloration, length of internode, pubescence density, and height; (4) leaf: attitude, type, length, width, type of blade, intensity of green color, glossiness, blistering, and anthocyanin coloration of veins; (5) inflorescence: type; (6) flower: color, corolla color, corolla blossom type, and pubescence of style; (7) peduncle and pedicel: abscission layer and length; (8) fruit: green shoulder before maturity, extent of green shoulder before maturity, intensity of green color of shoulder before maturity, intensity of green color excluding shoulder before maturity, green stripes before maturity, exterior color of immature fruit, pubescence, size, weight, length, ratio length/diameter, size homogeneity, shape in longitudinal section, ribbing at peduncle end, depression at peduncle end, size of peduncle scar, size of blossom scar, shape at blossom end, diameter of core in cross-section in relation to total diameter, thickness of pericarp, flesh color of pericarp, flesh color intensity, number of locules, color at maturity, intensity of exterior color, color of flesh at maturity, glossiness of skin, color of epidermis, color (intensity) of core, size of core, shape of pistil scar, fruit cross-sectional shape, easiness of fruit wall (skin) to be peeled, firmness, shelf-life, and time of maturity; and (9) seed: shape and color. All characters were recorded for four plants per replicate of each tomato line, for a total of 256 plants evaluated.

In our investigation, the following agronomic traits were evaluated: internode length (IL), stem length (SL), leaf length (LL), leaf width (LW), pedicel length (PL), equatorial diameter of the fruit (EDF), polar diameter of the fruit (PDF), polar/equatorial diameter ratio (EPR), number of locules (NL), shelf-life (ShL), and yield (Y).

2.3. Statistical Analyses

Using the information obtained from morphological characterization, the percentage of lines that presented similar characteristics was determined. Agronomic traits were analyzed using one-way ANOVA according to Block Complete Design, post hoc Tukey’s test (α = 0.05), and Pearson’s correlation coefficient. Canonical discriminant analysis was performed using lines with replicates as categorical variables and agronomic traits as dependent variables. Finally, the theoretical selection value for each line is calculated using the Eigenanalysis selection index method (ESIM) proposed by Cerón-Rojas et al. [16]. All analyses were performed using the Statistical Analysis System software version 9.1 [23].

3. Results

3.1. Morphological Characterization

Qualitative morphological trait data showed notable differences for 43 of the 60 traits evaluated for stem, leaf, and fruit characteristics. These differences were noted between the lines of the same type. In seedlings, anthocyanin pigmentation of the hypocotyl was observed in fourteen of sixteen lines; lines JCM-15 and JCM-17, both creole chino fruit shapes, did not exhibit this pigmentation. Regarding stem traits, internode length was short in 37.5% of the lines, intermediate in 62.5%, and long only in JCM-12. Pubescence density was scarce in 50% of the lines, intermediate in 43.75%, and dense only in JCM-06. Short plant height was observed in 13 lines, whereas JCM-01, JCM-03, and JCM-11 (all kidney fruit shapes) showed intermediate heights.

Regarding leaf characteristics, the predominant leaf attitude was horizontal in 12 lines (75%), whereas JCM-12 was drooping, JCM-10 and JCM-11 were semi-drooping, and JCM-08 was semi-erect. Leaf length was short in JCM-06 (cherry), medium in 45.75%, and long in 54.25% of the lines. JCM-06 (cherry) had narrow leaves, JCM-01 and JCM-17 had broad leaves, and the other lines had medium values. The intensity of the green color was mainly medium (68.75%), dark in JCM-12, JCM-15 and JCM-18, while JCM-06 and JCM-08 were light. The multiparous inflorescence type was predominant (81.25%), and uniparous was observed in JCM-06, JCM-09, and JCM-10. Pedicel length was short in 25%, medium in 43.75%, and long in 31.25% of the lines.

Regarding fruit characteristics, the extent of the green shoulder before maturity was small at 43.75%, medium at 31.25%, very small at JCM-01 and JCM-15, and large at JCM-08 (half kidney) and JCM-09 (saladette). The intensity of the green color of the shoulder before maturity was dark in 50% of the lines and medium in the rest, mainly in the kidney and creole chino lines. Twelve lines showed light intensity of the green color excluding the shoulder before maturity, while JCM-06 (cherry), JCM-07, JCM-11 (kidney), and JCM-18 (creole chino) showed medium intensity.

Green stripes before maturity were present in 56.25% of the lines, whereas JCM-05 (cherry), JCM-09 (saladette), JCM-10 (kidney), and the creole chino lines did not exhibit these stripes. The exterior color of immature fruit varied as follows: 43.75% of the lines were light green, 31.25% were green, and JCM-01, JCM-04, JCM-12 (kidney), and JCM-09 (saladette) were greenish white. Pubescence was intermediate in 87.5% of the lines and sparse in JCM-02 (kidney) and JCM-17 (creole chino).

Fruit size was medium in the kidney and half kidney lines, except for JCM10, which showed large fruits similar to all creole chino lines. Small fruit sizes were observed in JCM-09 (saladette) and JCM-11 (kidney), whereas JCM-06 (cherry) had very small fruit sizes. Size fruit homogeneity was observed in 75% of the lines, whereas JCM-08 (kidney), JCM-16, JCM-17, and JCM-18 (creole chino) showed intermediate homogeneity. Five shapes of the longitudinal fruit sections were recorded: 50% were oblate, while lines JCM-03, JCM-06 and JCM-09 were rounded, JCM-15 and JCM-16 were oblong, JCM-17 and JCM-18 were cordate, and JCM-08 was flattened.

Ribbing at the peduncle end was very strong in JCM-01, JCM-10 and JCM-12 (creole chino), strong in JCM-04, JCM-07 (kidney) and JCM-05 (cherry), medium in JCM-02 and JCM-08, weak in JCM-11, and weak in JCM-03, JCM-06, and JCM-09. The depression at the peduncle end was strong in JCM-01 and JCM-10 (kidney), medium in JCM-07 (kidney) and JCM-15 (creole chino), and weak in JCM-03, JCM-11 (both kidneys), JCM-5 (cherry), JCM-08 (half kidney), JCM-16, and JCM-17 (creole chino). None of the other lines had any depression.

The size of the peduncle scar varied among lines. JCM-02, JCM-03, JCM-12 (kidney), JCM-06 (cherry), and JCM-09 (saladette) had very small peduncular scars. JCM-04 and JCM-11 (kidney) displayed small peduncular scars, whereas JCM-01, JCM-07 (half kidney), JCM-15, JCM-17, and JCM-18 (creole chino) displayed medium peduncular scars. By contrast, JCM-08 (half kidney), JCM-05 (cherry), JCM-10 (kidney), and JCM-16 (creole chino) exhibited large peduncular scars.

The sizes of the blossom scars varied. JCM-02, JCM-03 (kidney), JCM-05, JCM-06 (cherry), and JCM-16 (creole chino) all exhibited very small blossom scars. JCM-04, JCM-11 (kidney), JCM-08 (half kidney), JCM-15, and JCM-18 (creole chino) showed small scars. JCM-07 and JCM-12 (kidneys) displayed medium scars, whereas JCM-01 and JCM-10 exhibited very large scars.

Four shapes were distinguished at the blossom end: flat (JCM-04, JCM-06, JCM-07, JCM-09, JCM-11, and JCM-16), indented to flat (JCM-01, JCM-02, JCM-03, JCM-05, and JCM-12), flat to pointed (JCM-08, JCM-16, JCM-17, and JCM-18), and pointed (JCM-10). The diameter core in the cross-section in relation to the total diameter of the fruit was large in 50% of the lines, very large in JCM-01, JCM-05, JCM06, JCM-10 and JCM-12, and medium in JCM09, JCM-16, and JCM-18.

The thickness of pericarp was medium in 43.75% of the lines, thin in all kidney lines and JCM05 (cherry), while it was very thin in JCM-06 (cherry), JCM-10, and JCM-11 (kidney). The flesh color of pericarp was red in 81.25% of the lines, and orange in JCM-06 (cherry), JCM-12 (kidney), and JCM-17 (creole chino). The red flesh color was very intense, whereas the orange color had an intermediate tone. The same values were observed for the color at maturity, intensity of exterior color, and color of flesh at maturity, although the glossiness of the skin was strong in red fruits and medium in orange fruits.

The color (intensity) of the core was light in all the lines. The size of core was large in 62.5% of the lines, very small in JCM-06 (cherry), small in JCM-09 (saladette) and JCM-11 (kidney), and intermediate in JCM-04, JCM-12 (kidney), and JCM-08 (half kidney). Three shapes of pistil scar were observed: 87.5% were dotted, JCM-07 (kidney) showed a stellate shape pistil scar, and JCM-12 (kidney) exhibited an irregular shape.

All lines displayed round shapes in the fruit cross-section, except for the kidney-shaped types, which showed irregular shapes. Easiness of fruit skin to be peeled was intermediate in 75% of the lines, whereas JCM-01, JCM-02, JCM-03, and JCM-10 (kidney) showed an ease of peeling. Intermediate firmness was observed in 62.5% of the lines and high firmness in JCM-11, JCM-12 (kidney), JCM-09 (saladette), JCM-17, and JCM-18 (creole chino).

All lines exhibited indeterminate growth habits, stems without pigmentation, standard-type leaves with bipinnate blade-type blades, weak glossiness, weak blistering, and lightly colored veins. Flowers were yellow with an open corolla, without pubescence on the style, and with an abscission layer. Fruits showed green shoulders before maturity, colorless epidermis, a shelf-life higher than five weeks, and flowering from 40 days after transplantation. The seeds were triangular in shape, with a pointed base, and were light yellow in color.

3.2. Agronomic Traits

All agronomic traits showed significant differences (p ≤ 0.05). Internode length (IL) was longest in JCM-12 (14.48 cm) and shortest in JCM-17 (8.39 cm). In kidney shape type, this trait varied between 9.96 and 14.48 cm, while creole chino varied between 8.39 and 13.88 cm, and cherry and saladette types averaged 9.5 cm (

Table 2. Mean tests of eleven agronomic traits evaluated in sixteen elite tomato lines and selection values by mean ESIM (Eigenvalue selection index method) coefficient.

Line	IL	SL	LL	LW	PL	EDF	PDF	EPR	NL	ShL	Y	ESIM
JCM-01	9.96 def	1.60 abc	11.75 ab	7.75 a	0.50 gh	5.95 b	3.70 defg	0.62 g	6.00 bc	6.00 a	1937.83 ab	629.60
JCM-02	10.91 cde	1.53 abcd	8.75 cde	6.38 abcd	0.48 gh	4.80 cdef	3.60 defg	0.75 defg	5.00 bcd	7.00 a	1325.73 abcde	431.20
JCM-03	12.47 abc	1.65 ab	8.75 cde	6.00 bcde	1.03 bcd	4.60 efg	3.38 efg	0.74 efg	4.50 def	6.50 a	1842.98 abc	597.00
JCM-04	11.07 cde	1.56 abc	7.25 ef	5.00 def	0.93 cde	4.73 defg	4.05 def	0.86 def	4.75 cde	6.00 a	1814.80 abc	587.50
JCM-05	9.47 ef	1.43 abcde	12.50 a	6.50 abcd	0.58 fgh	5.70 bcd	4.38 cde	0.77 defg	5.00 bcd	6.25 a	672.53 cdef	224.10
JCM-06	9.61 ef	0.92 g	5.50 f	4.13 f	0.33 h	1.55 i	1.50 h	0.97 bcd	2.00 i	6.50 a	94.90 f	33.47
JCM-07	12.68 abc	1.40 bcde	11.25 abcd	7.00 abcd	0.63 efgh	5.40 bcde	3.40 efg	0.63 g	6.25 b	5.50 a	542.75 def	181.50
JCM-08	10.75 cde	1.28 def	11.75 ab	5.75 cdef	1.08 bc	5.75 bc	5.23 bc	0.93 cde	4.50 def	6.75 a	1487.80 abcd	485.40
JCM-09	9.57 ef	1.37 cde	9.25 bcde	4.75 ef	0.83 cdef	2.73 gh	3.85 g	0.77 defg	2.25 hi	6.25 a	158.73 ef	56.84
JCM-10	11.10 cde	1.22 ef	12.50 a	6.75 abcd	0.75 defg	7.18 a	4.63 cd	0.64 fg	8.50 a	6.50 a	2404.95 ab	779.60
JCM-11	11.95 bcd	1.69 a	11.50 abc	6.25 abcde	0.88 cdef	3.48 h	2.60 gh	0.75 defg	3.75 defg	6.50 a	151.38 ef	55.29
JCM-12	14.48 a	1.38 bcde	8.50 de	6.00 bcde	0.43 h	4.93 cdef	2.93 fg	0.59 g	6.25 b	6.00 a	241.55 ef	83.09
JCM-15	10.63 cde	1.07 fg	12.75 a	6.50 abcd	1.25 ab	4.38 fgh	3.55 defg	0.81 defg	3.50 efgh	5.50 a	717.63 cdef	238.30
JCM-16	13.88 ab	1.19 efg	11.75 ab	7.25 abc	1.40 a	6.00 b	7.00 a	1.17 ab	3.00 ghi	5.75 a	1219.00 bcdef	401.00
JCM-17	8.39 f	1.28 def	12.00 ab	7.50 ab	0.58 fgh	5.35 bcdef	6.45 a	1.21 a	3.25 fghi	5.75 a	2499.83 a	811.40
JCM-18	9.27 ef	1.25 ef	9.50 bcde	6.50 abcd	1.00 bcd	5.50 bcde	6.15 ab	1.12 abc	3.00 ghi	5.75 a	1284.45 bcdef	420.40
HSD	2.06	0.27	2.92	1.7	0.3	1.02	1.13	0.23	1.5	1.67	1.207.30
CV	7.34	7.89	11.07	10.57	15.11	8.07	10.80	10.67	13.12	10.62	21.11
TS	Negative	Negative	Positive	Positive	Positive	Positive	Positive	Positive	Negative	Positive	Positive

Values accompanied by the same letter per column did not show significant differences according to Tukey’s test (α = 0.05); HSD: Tukey’s honest significant difference; CV: coefficient of variation; TS: type of selection to calculate the ESIM coefficient (negative means selecting lower values and positive means selecting higher values); IL: internode length (cm); SL: stem length (m); LL: leaf length (cm); LW: leaf width (cm); PL: pedicel length (cm); EDF: equatorial diameter of the fruit (cm); PDF: polar diameter of the fruit (cm); EPR: equatorial/polar diameter ratio; NL: number of locules; ShL: shelf-life (weeks); Y: yield (g·plant-1); ESIM: selection ESIM coefficient (higher values indicate that it is more selectable). Pedicel length (PL) varied between 0.33 (JCM-06) and 1.4 cm (JCM-16), where creole chino and kidney types showed values higher than 1 cm.

). Stem length (SL) was longest in JCM-11 (kidney) at 1.69 m and shortest in JCM-06 (cherry) at 0.92 m. All kidney and creole chino shape types showed values higher than 1.2 m.

Leaf width (LW) was 6 cm in 50% of the lines. JCM-01 (kidney) had the widest leaf (7.75 cm) and JCM-06 (cherry) had the shortest (4.13 cm). Leaf length (LL) was longest in JCM-05 (cherry) at 12.5 cm and shortest in JCM-06 (cherry) at 5.5 cm (Table 2).

The equatorial diameter of the fruit (EDF) showed high variability. JCM-10 (kidney) had the highest value (7.18 cm), whereas JCM-06 (cherry) had the lowest value (1.55 cm). For the polar diameter of fruit (PDF), JCM-16 (creole chino) had the highest value (7 cm) and JCM-06 (cherry) had the lowest value (1.5 cm). JCM-17 (creole chino) showed the highest polar/equatorial diameter ratio (EPR), whereas JCM-01, JCM-07, and JCM-12 (kidney) showed the lowest EPR (Table 2).

The number of locules (NL) was variable: a higher number (8.5) was observed in JCM-10 (kidney) and the lowest (2) in JCM-06 (cherry) and JCM-09 (saladette). In the creole chino lines, NL was constant, whereas in the kidney lines, NL was variable. All lines had a shelf-life higher than five weeks (Table 2). The highest yield per plant was observed in JCM-17 (creole chino) and the lowest in JCM-06 (cherry). The most selectable lines according to the ESIM coefficient were JCM-17 (creole chino), JCM-10 (kidney), and JCM-01 (kidney), which showed the best combinations of agronomic traits (Table 2).

3.3. Correlation Analysis

Significant positive correlations were found between Y with EDF and PDF. In addition, significant positive correlations were found between LL and LW with EDF and PDF. A highly significant negative correlation was found between EPR with IL. Finally, ShL showed no significant correlation (

Table 3. Pearson’s correlation coefficient among eleven agronomic traits evaluated in sixteen elite tomato lines.

	SL	LL	LW	PL	EDF	PDF	EPR	NL	ShL	Y
IL	0.1928	−0.05234	0.05845	0.2215	0.13311	−0.09054	−0.32283 **	0.28685 *	−0.10688	−0.1689
SL	1	0.00192	0.17419	−0.07093	0.15359	−0.10928	−0.3664 **	0.25521 *	0.08021	0.19422
LL		1	0.53664 ***	0.30992 *	0.57161 ***	0.41885 ***	−0.032	0.2511 *	−0.07884	0.24834 *
LW			1	0.06007	0.55348 ***	0.46625 ***	0.05347	0.34631 **	−0.22486	0.39237 **
PL				1	0.21423	0.44299 ***	0.32037 **	−0.25191 *	−0.18584	0.07987
EDF					1	0.67346 ***	−0.09553	0.59126 ***	−0.12046	0.57946 ***
PDF						1	0.65878 ***	−0.03728	−0.17261	0.50921 ***
EPR							1	−0.6073	−0.09726	0.13381
NL								1	0.02931	0.29035 *
ShL									1	0.07014

* p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001; IL: internode length (cm); SL: stem length (m); LL: leaf length (cm); LW: leaf width (cm); PL: pedicel length (cm); EDF: equatorial diameter of the fruit (cm); PDF: polar diameter of the fruit (cm); EPR: equatorial/polar diameter ratio; NL: number of locules; ShL: shelf-life (weeks); Y: yield.

).

3.4. Canonical Analyses

On the other hand, the canonical discriminant analysis showed that three canonical roots (CAN) were necessary to explain 80.61% of total variability. CAN1 was positively correlated with EDF and PDF (fruit size), while CAN2 was positively correlated with EPR and negatively correlated with NL. Finally, CAN3 was positively correlated with PL, IL, and SL (

Table 4. Eigenvalues and total canonical structure of sixteen tomato elite lines using eleven agronomical traits.

Trait		CAN1	CAN2	CAN3
Percentage of variance (%)		40.19	30.12	10.30
Cumulative variance (%)		40.19	70.31	80.61
Stem	Internodes length (IL)	0.126785	−0.363094	0.493917
	Length (SL)	0.102564	−0.516271	0.46796
Leaves	Length (LL)	0.679915	−0.029199	0.1164
	Width (LW)	0.691714	−0.081496	0.019515
Pedicel	Length (PL)	0.367118	0.493065	0.563502
Fruit	Equatorial diameter (EDF)	0.944592	−0.164329	−0.104123
	Polar diameter (PDF)	0.782022	0.536394	−0.001741
	Equatorial/polar diameters ratio (EPR)	0.087989	0.884716	−0.044594
	Number de locules (NL)	0.497645	−0.761367	−0.275491
	Shelf-life (ShL)	−0.174778	−0.180918	−0.053089
	Yield (Y)	0.592127	0.110843	−0.358257

).

In the three-dimensional projection, JCM-10 (kidney) had the highest positive value for CAN1, which indicated the largest fruit size, whereas JCM-06 (cherry) had the lowest value for CAN1, representing the smallest fruit size. The lines with positive CAN2 values showed wider fruits with fewer locules, whereas negative values showed the opposite. JCM-16 showed the highest values in CAN3 owing to its longer internodes, stem, and pedicel (Figure 1).

4. Discussion

The differences in stem, leaf, and fruit traits observed in the lines evaluated in this study showed high variability among tomato lines, confirming that quantitative leaf and fruit traits are useful for distinguishing tomato morphotypes and populations [24,25,26,27,28]. However, in these studies, no variation was observed in leaf, stem, or fruit traits within the same tomato shape type, as in our study, which is probably because of the progress obtained during the genetic improvement process of these lines. They also revealed that five qualitative traits could be determining factors for differentiating tomato types and lines.

As a result of the morphological characterization carried out on the 16 lines, it is clear that it is important to use the characteristics of pubescence density, green shoulder size, cross-sectional shape, ribbing at the peduncle end, and depression at the peduncle end in the varietal characterization of tomatoes, not only focusing on quantitative characteristics that were used in previous studies to distinguish genotypes, exotic lines, breeding lines [24,25,26,28,29,30], and creole and native varieties [15]. Quantitative characteristics [31,32] can be complemented with the five identified in our study as determinants that prove relevant. Consequently, they should be used as markers of the tomato types, which were the object of this study.

On the other hand, our results show that the five tomato types characterized have common traits in plant: growth habit; stem: pigmentation; leaf: type, type of blade, glossiness, blistering, and anthocyanin coloration of veins; flower: color, corolla color, corolla blossom type, and pubescence of style; peduncle: abscission layer; fruit: green shoulders before maturity, color of epidermis, shelf-life, and time of maturity; and seed: shape and color. They also show that each line has its own characteristics that differentiate them from each other and that are relevant in the selection processes [18,24,25,26,29,30]. However, they were not decisive in distinguishing the types of tomato shapes [24,30,33,34].

The results obtained from the 11 agronomical traits revealed differences even within the same fruit shape. Variations in the IL and ST were consistent with those reported in elite line evaluations [5]. The wide interval between LL and LW demonstrated differences in plant archetype structure. This can be observed in the cherry-type lines, which showed contrasting values. Similarly, differences in PL, mainly in the creole chino lines, and in some kidney-type lines, indicate the variability within each fruit type [15,31].

The wide diversity in fruit size reported in previous studies is consistent with our results. This can be seen in the values of PDF ranging from 1.88 to 9.57 in native varieties, EDP averaging 4.97 cm for elite lines, as well as 3.29 PDF and 3.81 EDF for creole accessions have been reported [5,31,35]. Values in PDF and EDF in this research noted typical variation in fruit shape type, where the cherry type was smaller and the creole chino type was larger. Moreover, kidney-type lines were more compact than other types, although all creole and native germplasms had an EPR close to 1, indicating rounded fruits [31,36].

Most of the collected germplasm, accessions, improved lines, and local varieties developed by creole or native germplasm show variable NL in fruits of the kidney fruit shape type and constant NL in creole chino [35], which is consistent with our results. The individual selection method used to obtain elite lines improved the shelf-life. Thus, the long shelf-life gives the native and creole germplasms a long period of conservation and commercialization in the different links of the value chain [37], which represents a significant advantage.

The performance shown by the lines with the creole chino-type fruit reveals the yield potential that can be achieved with the use of improved plants obtained from native germplasm [4,35], which can significantly exceed the yield obtained with commercial genotypes [28].

The agronomic traits evaluated in our study are valuable for use in genetic improvement programs to improve the characteristics of fruits, flowers, and leaves. They could be complemented with other morphological traits, such as the type of ramification, distal fruit shape, epicarp color, and seed shape, which were previously used in tomato characterization [36,38].

The use of selection indices in this study allowed us to identify five elite lines (JCM-01, JCM-03, JCM-04, JCM-10, and Y JCM-17) with relevant characteristics to start a genetic improvement process based on a good combination of highly desirable morphological traits (LL, LW, PL, DEF, PDF, EPR, ShL, and Y). The eight characteristics observed in our study with higher Eigenvalues located on the three axes influenced the determined diversity with values higher than 0.5, and they coincided with the characteristics of fruit shape, fruit width, number of locules determined in native collections from Mexico [31], and commercial genotypes [27]. In addition, it allows for the discrimination of the IL, SL, and NL traits by presenting negative values [39]. Thus, a new approach to the genetic improvement of this species is proposed.

Correlation analysis may play an essential role in predicting the performance of some positively correlated traits for another trait in indirect selection, and vice versa. Hence, one or more traits in the same group would be a reliable parameter for the selection of plants with desirable content of the other positively correlated traits [40], that is, in this case, the fruit size traits with yield and the leaf size traits with fruit size traits. In the same context, a trait that negatively correlates with other traits would also be a reliable indicator as a selection criterion; in our case, a lower value of IL and a higher value of EPR, and vice versa.

The estimation of the contribution of each line to the variability present in each axis of the three-dimensional projection allowed the identification of seven characteristics with the greatest positive distance [16,41]. In this way, it is explained that the variability is determined by the characteristics related to the morphology of the stem (IL and SL), pedicel (PL), and fruit (EDF, PDF, EPR, and NL) [38], which can be considered in the improvement programs.

5. Conclusions

Morphological characterization allowed us to detect variations in plant and fruit morphological traits among the elite lines evaluated. This revealed that pubescence, extent of the green shoulder before maturity, shape in the longitudinal section, ribbing at the peduncle end, and depression at the peduncle end were transcendental qualitative characteristics that denote the differences among the elite lines of native tomatoes. Regarding agronomic traits, elite lines showed good shelf-life and notable productive value, confirming the progress of breeding efforts in these lines. Fruit-related traits were particularly significant in the characterization and selection criteria of the lines, as evidenced by canonical discriminant analysis and ESIM values. Thus, the morphological variation found and the characteristics with higher selection values identified may be valuable for optimizing the tomato genetic improvement process in general.