Evaluation of purple-fleshed sweetpotato genotypes for root yield , quality and pest resistance

Production of purple-fleshed sweetpotatoes, source of anthocyanins, pigments with high antioxidant activities, is expected to be extended during the upcoming years, as there is a strong trend to market products promising improved health. This study aimed to evaluate yield, postharvest quality and pest resistance of roots of purple-fleshed sweetpotatoes. Six genotypes (BGBD 0005, BGBD 0080, BGBD 1261, BGBD 1399, BGBD 1402 and BGBD 1405) and the cultivars Beauregard and Brazlândia Roxa as controls were evaluated during two seasons 2018 (Experiment 1) and 2019 (Experiment 2), in Brasília-DF, Brazil. The experiments were conducted in a complete randomized blocks design with four replications, with experimental plots consisting of two rows of six plants. Roots were harvested about 140 days after planting and evaluated for yield, appearance, insect damage and quality characteristics as soluble solids, dry matter and color. All genotypes showed roots with good resistance to soil insects. The highest commercial production of genotypes BGBD 1261, BGBD 0005, and BGBD 1405 associated to quality traits (dry matter content and soluble solids) and pest resistance, showed their potential for being released as cultivars or to be used in breeding programs.

Palavras-chave: Ipomoea batatas, raízes comerciais, aparência, danos por insetos, sólidos solúveis totais, matéria seca. RAC Melo et al. high retail prices. Some vegetables and fruits, such as red onion, cabbage and table beet, açaí and black grapes can be considered exceptions, as they are widely distributed in the country. Thus purple-fleshed sweetpotatoes can become a year-round/feasible alternative to those species.
With proper management practices and adoption of technologies, yields can reach from 25 t ha -1 to ≥30 t ha -1 in 4 to 5 months (Andrade Júnior et al., 2012;Carmona et al., 2015;. The average national yield of sweetpotato of 14.5 t ha -1 (IBGE, 2019) is considered low, which can be attributed to the continuous use of slips or vines with systematic accumulation of diseases that cause their degeneration. Inadequate production systems and also low fertility soils result in genotypes/ cultivars not expressing their full genetic potential (Carmona et al., 2015, Melo et al., 2019. In Brazil, from the 29 sweetpotato cultivars registered in the RNC (National Cultivars Register), SCS370 Luiza is the only one with purple-flesh color (BRASIL, 2020), with restricted recommendation to Santa Catarina state and reported average yield of 14.7 t ha -1 (Schallenberger et al., 2017).
Therefore, the development of new purple-fleshed genotypes with higher root yield, quality and also pest resistance is needed. Many losses due to insect damages are reported, causing disposal or rejection by retailers and consumers (Ames et al., 1996;Edmunds et al., 2008;Moyer, 2018). The major pest species responsible for direct damages to sweetpotato roots in Brazil are Euscepes postfasciatus, Diabrotica speciosa, Diabrotica bivittula, Sternocolaspis quatuordecimcostata and Conoderus spp. (Gallo et al., 2002).
Chemical pesticides are not highly effective for controlling these pest species since they develop in the soil and are protected in the roots. Thus, their management should be focused on the cultural control and plant resistance. These management practices approaches combined with plant resistance of some genotypes (Collins et al., 1999) may be the most effective for pest control (Barreto et al., 2011;Andrade Junior et al., 2012;Massaroto et al., 2014).
Thus, in this study we aimed to evaluate yield, postharvest quality and pest resistance of roots of purple-fleshed sweetpotato genotypes.
On both experiments, the preplant fertilization included 120 kg ha -1 ammonium sulfate, 440 kg ha -1 single superphosphate, 12 kg ha -1 boric acid and 20 kg ha -1 of ammonium zinc. Slips/ vines were transplanted 3 inches deep with 4 plant nodes underground, and 2 to 3 nodes above the ground. Weed control by hand-hoeing and topdress N fertilization (50 kg ha -1 ) were performed 30 days after planting (DAP).
At 140 DAP (Exp. 1) and 138 DAP (Exp. 2), the center two rows of each plot were harvested and roots were evaluated for the following components of production: 1) number of commercial roots (CRN); 2) mass of commercial roots [CRM, (t ha -1 )]; 3) total number of roots (TRN); 4) total mass of roots [TRM (t ha -1 )]; 5) CRM/CRN ratio, in g. Appearance of roots (RA) was also evaluated, using visual index scores corresponding to: 1= non standards, with very irregular shape, large veins and cracks; 2= very nonuniform, with the presence of large veins and cracks; 3= nonuniform, with large veins and cracks; 4= slightly nonuniform with the presence of veins; and 5= regular fusiform without veins or cracks (Andrade Junior et al., 2012). Commercial roots were the ones with RA scores ranging from 4 to 5 and ID scores from 1 to 3, and weighing between 150 and 1500 g.
After sampling, 10 roots at random per plot, insects damage (ID) was evaluated with the index score as follow: 1= free of insect damage; 2= few damages; 3= few commercially damaged roots; 4= most commercially damaged roots; and 5= roots commercially unacceptable for both human and animal consumption (Massaroto et al., 2014).
The number of perforations (NP) per root, diameter (RD), length (RL) and skin thickness (ST), in mm, using a digital caliper, were also evaluated. Additionally, for the second experiment (Exp. 2), 10 roots were sampled randomly in each plot and evaluated for dry matter [DM (%)] and soluble solids (SS) content. Each genotype had its flesh crushed and the liquid portion was used to determine SS in a digital refractometer (PR-101, Atago Co. Ltda., Tokyo, Japan). Results were expressed in o Brix (AOAC, 2005). DM was determined by weighing the samples (5 g) before and after drying in an oven (Quimis, São Paulo-SP, Brazil) at 105 o C for 3 h. Petri dishes were cooled in a desiccator until room temperature and then weighted. This procedure was repeated until constant samples weight (Quimis, São Paulo-SP, Brazil). DM was calculated according to the following equation: DM% = dry weight (g)/fresh weight (g) x 100.
Data were tested for normal distribution by Lilliefors test and submitted to analysis of individual and combined variance for the two periods of experiments, and cluster means were compared by Scott-Knott grouping test at 5%. All statistical analyses were performed using Genes software (Cruz, 2013) and AgroEstat (Barbosa & Maldonado Júnior, 2015).

RESULTS AND DISCUSSION
The analysis of variance showed that there was an interaction between the two periods of experimentation and genotypes for CRN (commercial root number), CRM (commercial root mass), TRN (total root number), TRM (total root mass), CRM/CRN ratio, RA (root appearance), ID (insect damage), SC (skin color) and FC (flesh color). NP (number of perforations) showed a higher CV of 68.67%; however, its relation of genetic and phenotypic coefficient (CVg/CV) was greater than one, denoting a predominance of genetic variance, making possible to use this information to select the best genotypes (Table 1 and 2).
In Exp. 1, the BGBD 1261 genotype showed the highest CRM (51.11 t ha -1 ) followed by BGBD 0005 (36.25 t ha -1 ), BGBD 1402 (35.38 t ha -1 ), Beauregard (28.18 t ha -1 ) and Brazlândia Roxa (35.46 t ha -1 ). In Exp. 2, Beauregard had the highest CRM, (35.68 t ha -1 ). The higher CRM during Exp. 1 values may be associated with cooler temperatures and a better distribution of rainfall, presenting an average maximum temperature of 27.18°C, minimum of 17.63°C, air humidity of 73.2% and accumulated rainfall of 964.2 mm. The aforementioned weather conditions meet the requirements of sweetpotatoes and possibly lead to a good vine/ root ratio (Lebot, 2019). As for Exp. 2, an average maximum temperature of 21.8°C, minimum of 20.63°C and air humidity of 41.5%, associated to lower values of rainfall (accumulated 185 mm), with no registered values for June, July and only 10 mm during August, may possibly have impacted the obtained CRM.
The average CRM of the genotypes,  (Table 1). Appearance is considered the most influential sweetpotatoes attribute considered during their purchase decision-making process. Colored flesh cultivars are well accepted by consumers, as long as added to other attributes such as flavor and texture (Leksrisompong et al., 2012). Beauregard was less resistant to pest attack (predominantly Euscepes postfasciatus). This cultivar showed higher NP and ID scores than all   genotypes and Brazlândia Roxa ( Table  2). The fact that the purple-fleshed genotypes have grouped with Brazlândia Roxa (Table 2) may be an indicative of their resistance to pests since this cultivar is considered insect-resistant (Barreto et al., 2011;Andrade Junior et al., 2012;Massaroto et al., 2014;Amaro et al., 2019). R o o t s o f p u r p l e -f l e s h e d sweetpotatoes did not differ in length (RL) in both seasons (Table 2). Regarding the diameter (RD), in both 2018 and 2019 seasons, BGBD 0005, BGBD 1261 and BGBD 1402 showed the highest RD (Table 2). These genotypes showed no difference from Beauregard in 2018 and BGBD 0080 in 2019. Fusiformshaped sweetpotato is the most accepted for fresh commercialization (Andrade Junior et al., 2012). As the purplefleshed genotypes had similar lengths (mean value of 156.76 mm), those with a larger diameter (about 55 mm) ( Table 2) could be considered to have the best shape, with a ratio close to 1/3 of the diameter by length. Based on this standard, the best genotypes coincide with those with higher CRM (Table 1).
Skin thickness (ST), was similar for all genotypes, with a mean value of 2.49 mm ( Table 2). Roots of freshly harvested sweetpotato have a thin skin and therefore susceptible to mechanical damage (Edmunds et al., 2008). Thus, genotypes with thicker skin, which do not damage when washed and classified, are more recommended (Edmunds et al., 2008).
BGBD 1399 genotype showed an intense purple color and best scored for SC in Exp. 2, and in Exp. 1 this genotype was similar to BGBD 0005, BGBD 0080 and BGBD 1261 ( Table 2). The genotypes BGBD 1399 and BGBD 1405 showed the highest score for FC in Exp. 2 (Table 2), which may be an indicative of higher anthocyanin content. All other genotypes showed FC scores classified as light purple.
BGBD 1399 and BGBD 1405 showed 35.42% and 35.07% dry matter (DM), respectively (Table 2). These contents were higher than those found for Beauregard (22.19%) and lower than those of Brazlândia Roxa (39.48%). DM is an important quality attribute, especially for the production of flour, chips and other processed products, to maximize yield and to minimize oil or drying costs. Some cultivars adopted by the industry have DM around 30% (Mosta et al., 2015;Marangoni Junior, 2017). BGBD 1405 genotype also showed the highest SS contents, followed by BGBD 0080, BGBD 1261, BGBD 1399, and Beauregard (Table 2). Soluble solids mainly express the sugar content and may be related to a better flavor (Chitarra & Chitarra, 2005) and the contents found for the genotypes correspond to those found for other cultivars (Mu et al., 2017;Vizzotto et al., 2017;Sugri et al., 2019). SS is also an important quality character related to consumer's preferences. The six purplefleshed genotypes evaluated showed similar SS values as 10 genotypes studied by Oliveira et al. (2019). BGBD 1405 with 11.90 o Brix, value in close proximity to a purple-fleshed genotype named UGA-34, showed 13 o Brix in the aforementioned study, an indication of higher sucrose content, which can increase its acceptability.
The roots of all evaluated genotypes showed good levels of pest resistance, similarly to other purple-fleshed cultivars, such as Okinawan, which was described in Hawaii as having some resistance (Miyasaka et al., 2019).
This result is an important indicator of the substantial root commercial production of genotypes BGBD 1261, BGBD 0005, and BGBD 1405 associated to quality traits (dry matter content and soluble solids) and pest resistance, which showed their potential for being released as cultivars or to be used in breeding programs.