QUALITY OF VIRUS-FREE GARLIC GROWN UNDER HIGH ALTITUDE CONDITIONS IN THE SEMIARID REGION OF THE NORTHEAST OF BRAZIL

Garlic is a vegetable rich in starch and in aromatic substances; it has high value as a spice and is consumed as a fresh vegetable or in processed products. The garlic market demand has been growing in the last years, and virus-free garlic seed is a technology that can result in better quality products. The objective of this work was to evaluate the quality of virus-free garlic cultivars (VFGC) grown under high altitude conditions in the Semiarid region of the Northeast of Brazil. Three experiments were conducted in 2018 in Portalegre, state of Rio Grande do Norte, Brazil: the first experiment with first-generation VFGC (G1), the second with second-generation VFGC (G2), and the third with infected cultivars (without clonal cleaning). The experiments were conducted in randomized block experimental design, with four replications. The treatments consisted of five garlic cultivars (Amarante, Branco-Mossoró, Cateto-Roxo, Gravatá, and Hozan). The use of virusfree garlic seeds results in bulbs with larger diameter and higher soluble solid and total solid contents, pungency, and industrial yield, showing that the seed health is important for the quality of garlics for fresh consumption and industrial processing. The cultivars Amarante, Cateto-Roxo, and Hozan had the highest total solid contents; and the cultivar Hozan had the largest bulb diameter and the highest soluble solid contents, pungency, and industrial index. The cultivar Hozan is the most indicated for dehydration due to its higher industrial index. Bulb diameter, pungency, and industrial index presented positive correlations with soluble solids, pungency, and industrial index.


INTRODUCTION
Garlic is a vegetable that stands out as one of the main spices used for cooking in the world; it has good flavor and mainly aroma characteristics. Organosulfur compounds are responsible for pungency, flavor, and aroma characteristics of this vegetable (BESSA et al., 2017).
Most garlic in Brazil is marketed fresh, but the searching for ready products in the last years has increased the marketing of garlic in paste, cream, and in dehydrated, dried powdered, preserved, and encapsulated oil forms (LUCENA et al., 2016).
Considering this increasing demand of the consumer market, the use of adequate cultural practices for garlic crops is essential, as well the use of technologies such as the virus-free garlic seeds, which results in high-vigor plants, high bulb yields, and high-quality bulbs regarding health and physiology, and makes the product to reach the market with good quality either for fresh consumption or industrial processing (RESENDE et al., 2013).
Soluble solid, total solid, and total soluble sugar contents, pungency, titratable acidity, and industrial index are among the physical-chemical characteristics used to evaluate the garlic quality (LOPES et al., 2016). The pungency and total solid contents are used to evaluate these characteristics and indicate whether a product is more appropriate for fresh consumption or industrial processing (CHITARRA; CHITARRA, 2005).
The garlic chemical composition and the its sensorial characteristics flavor and odor depend more on the genetic factor (cultivar) than on the crop conditions; however, they are also dependent on edaphoclimatic conditions of the environment in which the garlic plants will be grown and developed (LIMA et al., 2019).
Some studies evaluated the production and quality of garlic of the groups semi-nobre (SOARES, 2013;BESSA et al., 2017;LUCENA et al., 2016) and nobre (LOPES et al., 2016;HENRIQUES et al., 2019;SILVA, 2019) grown in the state of Rio Grande do Norte, Brazil; however, only one reported the quality of the garlic from virus-free seeds (cultivar Roxo Pérola do Caçador) in this region, evaluating clove-seed size and spacing between plants (LIMA et al., 2019).
Thus, the objective of this work was to evaluate the quality of virus-free garlic cultivars grown under high altitude conditions in the Semiarid region of the Northeast of Brazil.

MATERIAL AND METHODS
Three experiments were conducted simultaneously from May to October 2018, in the municipality of Portalegre, state of Rio Grande do Norte (RN), Brazil (6°1'20"S, 38°1'45"W, altitude of 520 m). The climate of region is Aw, tropical rainy with a dry winter and a rainy summer, according to the Köppen classification, with mean annual rainfall depths from 800 to 1,200 mm (SEPLAN, 2019).
The first experiment was conducted with first generation of virus-free garlic cultivars (VFGC) (G1), the second with second-generation VFGC (G2), and the third with infected cultivars (without clonal cleaning). The experiments were conducted in a randomized block experimental design, with four replications.
The treatments consisted of five cultivars (Amarante, Branco-Mossoró, Cateto-Roxo, Gravatá, and Hozan) from the Brazilian Agricultural Research Corporation (Embrapa Vegetables, Brasília, DF, Brazil). The VFGC passed through a clonal cleaning process and were multiplied in controlled conditions.
The experimental plots consisted of beds of 0.20 m height, 1.0 m width, and 2.0 m length, with five planting rows. The seeds were planted to a depth of 0.05 m, with spacing of 0.20 m between rows and 0.10 m between plants, to a population of 100 plants in a total area of 2.0 m². The evaluated area of each plot consisted of the three central rows, discarding one plant from each end of the rows, resulting in a population of 54 plants evaluated in an area of 1.08 m 2 .
The bulbs were threshed, and the cloves were classified by size and treated with 2.5% iprodione solution before planting to prevent possible attack of soil pathogens.
Topdressing consisted of two applications, at 30 days after planting (DAP), using 30 kg ha -1 of N (urea); and at 60 DAP, using 30 kg ha -1 of N (urea) and 13 kg ha -1 of K 2 O (potassium nitrate). Plant health protection was carried out using products based on Mancozeb (Manzate WG ® ) and azoxystrobin (Amistar WG ® ) for purple spot, and a product based on chlorfenapyr (Pirate ® ) for pests, mainly thrips and mites. Weed control was carried out with manual weeding as needed.
A micro sprinkler irrigation system was used, with flow of 40 L h -1 and pressure of 200 KPa. The irrigation was suspended at five days before the harvest, when the first signs of plant maturation were observed leaf yellowing and drying of 2/3 of the shoots.
The plants were harvested manually and subjected to a pre-cure process, remaining for three days exposed to the Sun, and then cured for 17 days in a dry, shaded, airy place.
The bulbs were cleaned and samples of 10 bulbs per plot were evaluated for transversal diameter (mm) and mean bulb diameter (mm). The bulbs were then threshed and the cloves of these samples were manually peeled and crushed in a food processor to obtain a garlic pasta, which was evaluated for the following characteristics: soluble solids (SS): a juice obtained from the garlic pasta was homogenized, filtered in a 100% polyester fabric cloth, and read in a digital refractometer (Palette PR-100; Atago, Fukaya-shi, Japan). The results were expressed as percentages (AOAC, 2002).
total solids (TS): the cloves of ten bulbs were crushed and taken to a forced air circulation oven at 65 °C until constant weight. The total solids were calculated by the difference between 100 and the moisture content of the cloves. The results were expressed as grams of total solids per 100 grams of garlic (ZENEBON; PASCUET, 2005).
titratable acidity (TA): 1 g of garlic pasta from the crushed cloves was diluted in distilled water up to the volume of 50 mL. Two drops of 1% alcoholic phenolphthalein were added and the titration was carried out with a NaOH 0.1N solution until the turning point, characterized by a pink color. The results were expressed as mEq H 3 O + 100 g -1 (ZENEBON; PASCUET, 2005). pungency: estimated by determining the pyruvic acid, using a 2,4-dinitrophenylhydrazine (DNPH) reagent, and the colorimeter method described by Schwimmer and Weston (1961): an extract was obtained by adding 0.2 mL of the juice obtained from the garlic pasta, 1.5 mL of 5% trichloroacetic acid, and 18.3 mL of distilled water to a Erlenmeyer, which was shaken. An aliquot of 1.0 mL of the extract was added to a test tube, together with 1.0 mL of DNPH solution and 1.0 mL of distilled water. The solution was shaken in a vortex; the test tubes were then taken to a water bath at 37 °C for 10 minutes. The solution was cooled in an ice bath and 5mL of NaOH 0.6N was added per test tube. They were then shaken in a vortex and left to rest for five minutes to develop a yellow color. The absorbances were read in a spectrophotometer at 420 nm. Sodium pyruvate was used as standard. The pungency was calculated using the standard curve of sodium pyruvate with seven concentrations (0.0, 0.2, 0.4, 0.6, 0.8, 1.0, and 1.2 mmol L -1 ). The results were expressed in μMoles of pyruvic acid per mL of garlic juice. industrial index (II), calculated by the formula: II = (total solids × pyruvic acid) × 100, according to Carvalho et al. (1991).
The data of each experiment were subjected to analyses of variance. After meeting the assumptions of homogeneity of variances and normality of errors between the experiments, the data were subjected to joint analyses of variance; the means referring to the health of garlic seed and cultivars were compared by the Scott-Knott test (p≤0.05), using the Sisvar 5.4 program (FERREIRA, 2011). Pearson's correlation analysis was applied, using the t test at 1% and 5% probability.

RESULTS AND DISCUSSION
The interaction between garlic-seed health and cultivars was significant for all characteristics evaluated.
Plants from virus-free garlic seeds of first generation (G1) and second generation (G2) presented larger bulb diameter than the infected cultivars (Table 1).
The virus-free garlic cultivars (VFGC) presented larger bulb diameter than the infected cultivars; this is explained by the larger leaf area (plant height and number of leaves -data not presented) of VFGC in the field, which resulted in higher production of photoassimilates that promote the growth of bulbs. Leaves are essential for the photosynthetic process, which is required for synthesis of carbohydrates for the plant maintenance and growth and development of reserve organs (SINGH et al., 2013).
These results are consistent with those found in other studies (BIESDORF et al., 2015;MARODIN;SOUZA, 2018;LIMA et al., 2019), that showed larger bulb diameter for VFGC when compared to infected cultivars.
The G1 cultivars showed no significant differences in bulb diameter; however, considering the G2 cultivars, Branco-Mossoró, Cateto-Roxo, and Hozan presented higher means than Amarante and Gravatá. Among the infected cultivars, Hozan had larger bulb diameter (46.01 mm) than the other cultivars which had similar bulb diameters to each other (Table 1). Thus, Hozan had larger bulb diameter than the other cultivars, regardless of seed health.  The mean contents of soluble solid (SS) ( Table 2) and total solid (TS) ( Table 3) were higher for G1 when compared to G2 and infected cultivars. 1 Means followed by the same uppercase letter in the rows or lowercase in the columns are not different by the Scott-Knott test at 5% probability. These results were probably due to decreases in net photosynthetic rates of infected cultivars, which was followed by accumulation of sugar starch in the leaves, which negatively affected the translocation of photoassimilates to the bulbs and resulted in lower SS and TS when compared to the virus-free garlic plants. In addition, Basso et al. (2010) reported that viruses are concentrated in the phloem and can disorganize tissues conductors, affecting their normal functioning, resulting in accumulation of starch and sugars in the leaves, which can cause significant decreases in photosynthetic rates and source-drain relations.
Considering the G1 cultivars, Branco-Mossoró presented significant lower SS mean than the other cultivars, which had similar SS to each other. The G2 cultivars Cateto-Roxo and Hozan had higher SS percentages, followed by Amarante, Branco-Mossoró, and Gravatá, which had similar SS to each other. Among infected cultivars, Hozan had better performance, presenting higher SS (Table 2).
Differences in SS between cultivars may be caused by seed genetic characteristics and crop environment (LIMA et al., 2019). Mota et al. (2003) found 34% and 37.16% SS in the infected cultivars Cateto-Roxo and Amarante, respectively, which are similar results to those obtained in the present work.
The G1 cultivars presented higher total solid (TS) contents, with mean of 40.07%, except Branco-Mossoró (36.85%). Considering the G2 cultivars, Cateto-Roxo (40.67%) and Hozan (40.14%) had the higher TS means when compared to the other cultivars, which presented mean of 34.89%. The infected cultivars presented no significant difference in TS, presenting a mean of 35.94% (Table 3). TS above 30% are high and acceptable for processing, resulting in higher yields; thus, a high TS is a desirable characteristic for garlic industrial processing (CHAGAS; RESENDE; PEREIRA, 2003;BESSA et al., 2017). SS contents of VFGC were higher than those of infected cultivars. The garlic cultivars presented TS above 30%, with good prospects for industrial processing, regardless of their seed health.
The cultivars Amarante, Gravatá, and Hozan presented no differences in titratable acidity (TA) in the three experiments (G1, G2, and infected cultivars); however, The cultivar Branco-Mossoró presented the highest TA among the G1 cultivars, whereas Cateto-Roxo presented the highest among the infected cultivars (Table 4).
Branco-Mossoró in G1 and G2 had higher TA mean (7.53 and 6.79 mEq H 3 O + 100 g -1 , respectively) than the other cultivars. The infected cultivars presented no significant differences in TA (Table 4). 1 Means followed by the same uppercase letter in the rows or lowercase in the columns are not different by the Scott-Knott test at 5% probability.
Acidity in vegetable products is attributed to organic acids dissolved in the cell vacuoles in free and combined forms, such as salts, esters, and glycosides. TA is an important factor mainly for the industry, since the higher the acidity, the better the characteristics of garlic for use in industry (CHAGAS; PEREIRA, 2003;CHITARRA;CHITARRA, 2005). In this case, the cultivar Branco-Mossoró stands out with higher TA (6.81 mEq H 3 O + 100 g -1 ). Lopes et al. (2016) and Lima et al. (2019) pointed out that variations in TA are related to environmental factors and cultivar genetic characteristics. Chagas, Resende and Pereira (2003) evaluated infected garlic cultivars (without clonal cleaning) in Lavras, MG, Brazil, and found TA of 0.71% for the cultivar Amarante. Lucena et al. (2016) evaluated the cultivar Hozan in two locations and found low variations in TA between the locations: Governador Dix-sept Rosado (6.54) and Baraúna (6.79%). In the present study, the TA found for the cultivar Hozan and Amarante was 5.60% and 6.15%, respectively. Similar results found in different locations using the same cultivars confirm that TA is dependent on environmental factors.
The pungency found for the VFGC in G2, except Cateto-Roxo, was higher than that found for the VFGC in G1 and infected cultivars (Table 5). Lima et al. (2019), evaluated garlic plants of the Caçador cultivar and found higher pungency for virus-free seeds (85.97 μMols of pyruvic acid mL -1 ) when compared to infected cultivars (78.29 μMols of pyruvic acid mL -1 ). 1 Means followed by the same uppercase letter in the rows or lowercase in the columns are not different by the Scott-Knott test at 5% probability.
Among the G1 cultivars, Branco-Mossoró had the lowest mean pungency (43.17 μMols of pyruvic acid mL -1 ), and the other cultivars had mean of 74.0 μMols of pyruvic acid mL -1 , with no difference from each other. Among the G2 cultivars, Branco-Mossoró presented the highest pungency (98.46 μMols of pyruvic acid mL -1 ) and Cateto-Roxo the lowest (58.60 μMols of pyruvic acid mL -1 ); the other cultivars had mean of 83.13 μMols of pyruvic acid mL -1 . Branco-Mossoró and and Gravatá had higher pungency (76.09 and 80.89 μMols of pyruvic acid mL -1 , respectively) than the other infected cultivars, which presented mean of 69.38 μMols of pyruvic acid mL -1 (Table 5).
Other works developed in the same region also reported pungency in the infected cultivars Amarante, Branco-Mossoró, Cateto-Roxo, and Gravatá. Lucena et al. (2016) found pungency of 80.69% and 75.85 μMols of pyruvic acid mL -1 in plants of the cultivar Hozan, respectively, grown in Baraúna (120 m altitude) and Governador Dix-sept Rosado (34 m altitude). Bessa et al. (2017) found pungency of 64.57 μMols of pyruvic acid mL -1 in this cultivar, in Portalegre (520 m altitude). Soares (2013) found means of 94.84, 90.87, 94.46 and 92.58 μMols of pyruvic acid mL -1 in the cultivars Amarante, Cateto-Roxo, Branco-Mossoró and Gravatá, respectively. The results obtained in the present work and those of studies in other locations denote differences in pyruvic acid contents between cultivars in a same crop location, and for the same cultivar in different crop locations. Thus, regions with lower altitudes tend to present higher pungency. This was reported by Vargas et al. (2010)  Pungency is an important characteristic for the choice of cultivars, since the higher the pungency, the stronger the flavor and aroma of the finished product, which is desired by consumers (LUCENA et al., 2016;LOPES et al., 2016) for fresh consumption and industrial processing.
The cultivar Amarante presented higher mean industrial index (II) in G1 and G2 than in the experiment with infected cultivars; Branco-Mossoró and Hozan presented higher II in G2 than in the G1 and infected experiments; the cultivar Gravatá in G1, G2, and infected experiments presented no difference in II; and the cultivar Cateto-Roxo presented higher II in G1 than in the G2 and infected experiments (Table 6). Thus, the infected cultivars, except Gravatá, had lower II, indicating that seed health is important to obtain a high-quality industrial product. Lima et al. (2019) (Table 6). The industrial index is obtained by the correlation between total solid (TS) contents and pungency (μMols of pyruvic acid mL -1 ); therefore, cultivars that had higher TS and pungency had higher industrial index, with better quality for dehydration processing.
Several works have shown that II is a qualitative characteristic that varies depending on the environment in which the cultivars are grown. Lucena et al. (2016) found II of 26.42 and 27.22 for the cultivar Hozan, respectively, in Governador Dixsept Rosado and Baraúna. Soares (2013) found II of 27.00, 32.37, 30.22, and 29.04, respectively, for the cultivars Amarante, Branco-Mossoró, Cateto-Roxo, and Gravatá, in Governador Dix-sept Rosado. The II found in these studies are higher than those found in the present work and by Chagas, Resende and Pereira (2003), who found II of 19.42 for the cultivar Amarante, in Lavras, MG, Brazil. These results obtained in different locations denote that higher pungency and, consequently, industrial index is obtained in regions with low altitudes (VARGAS et al., 2010).
The correlations between bulb diameter and pungency (rF=0.99) and between bulb diameter and II (rF=0.71) were significant and positive, denoting that bulb diameter is favored by pungency and II; the correlation between pungency and II was also significant (rF=0.76) ( Table 7). 1 Means followed by the same uppercase letter in the rows or lowercase in the columns are not different by the Scott-Knott test at 5% probability. 1 ** = significant at 1%, and * = significant at 5% by the t test.
The correlation between SS and TS (rF=0.89) and between SS and II (rF=0.60) were significant, showing that SS contents are important for the quality of garlic for industrial processing, since the correlation between ST and II was highly significant (rF=0.86). Positive correlations denote simultaneous variance (covariance) of two variables in the same direction, i.e., proportional growth.

CONCLUSIONS
The use of virus-free garlic seeds results in bulbs with larger diameter and higher soluble solid and total solid contents, pungency, and industrial yield, showing that the seed health is important for the quality of garlics for fresh consumption and industrial processing.
The cultivars Amarante, Cateto-Roxo, and Hozan had the highest total solids contents, and the cultivar Hozan had the largest bulb diameter and highest soluble solid contents, pungency, and industrial index.
The cultivar Hozan is the most indicated for dehydration due to its higher industrial index. Bulb diameter, pungency, and industrial index presented positive correlations to soluble solids, pungency, and industrial index.