Yield and quality of curly kale grown using organic fertilizers

Kale is a vegetable that has high nutrition content, and balanced fertilization is essential to ensure high yield and quality of agricultural products. Curly kale, less known than regular leaf kale, is a new possibility for consumption. However, little is known about the influence of the type of fertilization on nutritional characteristics. This study aims to evaluate the influence of using organic fertilizers on the productivity, microbiological, and physico-chemical characteristics of Darkibor hybrid curly kale. The treatments consisted of three sources of organic fertilizers, one of organomineral fertilizer, and the control (without fertilization). The highest productivity was observed for kale that was treated with fertilizer in the organomineral composition in all harvests. For microbiological analyses, there was no contamination in all treatments, following legislation. There was no significant difference between treatments for the physicochemical composition, highlighting the high levels of phenolic compounds and proteins in curly kale. There was no difference between treatments regarding the mineral content of the leaves. Organic and organomineral fertilizers made it possible to produce curly kale with adequate


Research
Horticultura Brasileira 39 (1) January -March, 2021 B rassica plants are rich in nutrients a n d b i o a c t i v e c o m p o u n d s (Baenas et al., 2012), highlighting proteins (Lorenz & Maynard, 1988), carbohydrates, calcium, iron, iodine, vitamin A, niacin, vitamin C (Trani et al., 2015) and antioxidants (Baenas et al., 2012). Within this genus, species of B. oleracea evolved in many plant varieties where popular vegetables such as kale, brussels sprouts, cabbage, and broccoli are found essential sources of nutrients in many countries (Ordás & Cartea, 2008).
Kale is widely consumed in Brazil, especially collard, but curly kale is still little known in the consumer market. Unlike collard, curly kale, has curly leaves and is generally dark green (Olsen et al., 2009;Pathirana et al., 2017;Bejo, 2019). Also, it is high yielding and has good tolerance to early weighing (Bejo, 2019).

ABSTRACT
Kale is a vegetable that has high nutrition content, and balanced fertilization is essential to ensure high yield and quality of agricultural products. Curly kale, less known than regular leaf kale, is a new possibility for consumption. However, little is known about the influence of the type of fertilization on nutritional characteristics. This study aims to evaluate the influence of using organic fertilizers on the productivity, microbiological, and physico-chemical characteristics of Darkibor hybrid curly kale. The treatments consisted of three sources of organic fertilizers, one of organomineral fertilizer, and the control (without fertilization). The highest productivity was observed for kale that was treated with fertilizer in the organomineral composition in all harvests. For microbiological analyses, there was no contamination in all treatments, following legislation. There was no significant difference between treatments for the physicochemical composition, highlighting the high levels of phenolic compounds and proteins in curly kale. There was no difference between treatments regarding the mineral content of the leaves. Organic and organomineral fertilizers made it possible to produce curly kale with adequate physicochemical composition, free from microbiological contamination and heavy metals.
Vegetables are characterized by short growth cycles and high nutritional requirements. Vegetables of the genus Brassica have a high demand for calcium, sulfur, and boron (Filgueira, 2008), and for this reason, the supply of nutrients in a balanced way is essential to ensure high yield and quality of the vegetable. Studies have shown that organic fertilization has a positive result on the growth of these vegetables. Organic fertilizer is a product of plant, animal, or agro-industrial origin, which, when applied to the soil, improves fertility and increases crop productivity and product quality (Trani et al., 2013).
Organic fertilizers are generally found in solid and liquid form. Solid fertilizers are often incorporated into the soil before planting, and these organic materials are rich in slow-releasing organic N; it is difficult to predict the rate of mineralization in planning to meet crop uptake needs (Yoder & Davis, 2020).
According to , simple, mixed, compound, and organomineral organic fertilizers are classified according to their production's raw materials. Based on this normative instruction, the organic fertilizers in this research are referred to Class "A" and "B" fertilizers.
Class "A" organic fertilizer uses raw material generated from extractive, agricultural, industrial, agro-industrial and commercial activities, including those of mineral, vegetable, animal, industrial and agro-industrial sludge from wastewater treatment plants (authorized by the environmental agency), fruit and vegetable waste and food waste generated in pre-and post-consumption. These fertilizers are segregated at the generating source and collected by selective collection, free from waste or health contaminants, resulting in a product safe to use in agriculture. Class "B" fertilizer uses any amount of organic raw materials generated in urban, industrial, and agro-industrial activities in its production. It includes the organic fraction of urban solid waste from conventional collection, sludge generated in sewage treatment plants, industrial and agro-industrial sludge generated in wastewater treatment plants containing any amount of waste or health contaminants. These fertilizers are authorized by the environmental agency, resulting in a product safely used in agriculture.
Massive production of organic waste provides Brazil with a high potential for organic fertilizers or organic mineral fertilizers (Sá et al., 2017;Correa et al., 2018). Organomineral fertilizers can replace mineral fertilizers; they have equivalent performance and add organic compounds to the soil, potentially improving their properties and agricultural production (Mumbach et al., 2020).
The type of fertilizer used in the crop can influence the chemical composition of vegetables, affecting their biological quality (Bernardi et al., 2005). Knowledge of the effect of fertilizers on the characteristics of the vegetable can help expand the options of use and reduce production costs, especially considering that there is less knowledge about curly kale than the other types. In this context, this study aims to evaluate the effect of using three organic fertilizers (two from class A and one from class B) and one organomineral (class B) on the productivity, microbiological and physicochemical characteristics of curly kale. The work's hypothesis is replacing organomineral fertilizer with organic fertilizer without impairing production, microbiological health, and the physicochemical composition of curly kale.

Site and characteristics of the experiment
The experiment was carried out at the Center for Agricultural Sciences, at the Federal University of São Carlos (CCA/ UFSCar), Araras-SP (22º21'25"S, 47º23'03"W, 646 m altitude), a region characterized by dry winters and rainy summers.
The curly kale used was Brassica oleracea var. acephala, Darkibor hybrid. The seedlings were produced in trays with commercial substrate, and 30 days after germination, they were transplanted to the experimental plots in the field and irrigated using a sprinkler. The experimental plots were 12 m 2 with 50 x 50 cm spacing between plants and rows. The useful area of the plot used for the evaluations was ten central plants. The experimental design was of randomized blocks with five treatments (fertilizers and control) in four replications.
The evaluated treatments were: A) control, without fertilizer application; B) organomineral fertilization, class B [organic compound of materials for reuse and recycling of waste generated by human activities, macro and mineral micronutrients (0.7 kg/plot in planting and 0.05 kg/plot topdressed)]; C) organic fertilization 1, class A (comprising poultry litter, bone meal, industrial slaughterhouse waste, and food and other industrial waste (2.3 kg/plot in planting and 0.2 kg/plot topdressed); D) organic fertilization 2, class B [comprising sewage sludge (4.7 kg/plot in planting and 0.31 kg/plot topdressed)] and E) organic fertilization 3, class A [consisting of sugarcane filter cake (2.3 kg/plot in planting and 0.34 kg/ plot topdressed)]. Information on the description of the fertilizers was the same as that described on the product label.
The quantities of fertilizers used were defined according to the soil's chemical characteristics and based on the requirements of the NPK in planting and cover fertilization (Trani et al., 2015). The planting fertilizations were carried out one week before planting and immediately after the cuts, and the covering was carried out fortnightly until the 60 th day, and, after that date, it was carried out monthly.

Yield assessment
Kale productivity was evaluated in four harvests carried out at 60, 90, 120, and 150 days after planting. All the leaves of the plot's four central plants were collected, determining the mass of fresh matter by weighing.

Microbiological analyses
Microbiological analyses were performed on plants collected at 60, 90, and 120 days after planting. Initially, kale leaves were washed in running water, cut, and then immersed for 10 minutes in water with 200 mg/L sodium hypochlorite and washed again with sterile distilled water. The kale leaves were packed in sterile plastic bags, and 1 mL of 0.1% sterile peptone water was added per gram of sample. Each sample was washed by shaking the plastic bag vigorously 50 times carefully to avoid puncturing it or liquid overflow. This procedure was based on the methodology described in Paiva (2011), with modifications.
From the liquid remaining after washing the leaves, aliquots were removed to detect Salmonella spp. using the 1-2 Test kit (BioControl) after preenrichment of the samples in lactated broth at 35 o C for 24 hours. The total coliforms and Escherichia coli analysis was performed on Petrifilm TM EC plates (3M).

Physicochemical analysis
For the physicochemical analysis, the third or fourth fully developed leaf was selected, from the apex to the plant's base (Trani et al., 2015) of the four central plants.
The mass loss, turgor pressure, and instrumental color analyses were performed in the first harvest (60 days), at 0, 5, and 9 days of storage. Three packs (two leaves each) were placed in polyethylene bags (0.30 x 0.40 m) opened and stored in a refrigerated chamber at 10±1ºC and relative humidity between 85-90%.
Mass loss was determined by weighing, considering the difference between the initial and final mass. The Wiltmeter® equipment was used to determine the turgor pressure loss . The instrumental color was evaluated using the Hunterlab colorimeter, and the following was recorded: L value (lightness), which varies from black (L = 0) to white (L = 100); the a* value, which varies from red (+a*) to green (-a*); and the b* value that varies from yellow (+b*) to blue (-b*).
For the analysis of total soluble solids, pH, moisture, fibers, ash, phenolic compounds, minerals, and heavy metals were analyzed at 60 (1 st harvest), 90 (2 nd harvest), and 120 (3 rd harvest) days after planting, except for phenolic compounds, which were determined in the first two harvests. The total soluble solids ( o Brix) were analyzed by directly reading the supernatant in a digital bench refractometer (Atago model RX-5000α-Plus). The pH was determined in the homogenized material using an Edutec bench model pot EEQ9003-110 at 20ºC.
The fiber content was analyzed using the ANKOM method (2017). The moisture content was determined in an oven at 100-105°C until constant mass; the ash content was obtained by the burning method at 600-650°C (AOAC, 2012). The total phenolic compounds (gallic acid equivalent per 100 g) were determined according to the Folin-Ciocalteau spectrophotometric method with modifications (Singleton & Rossi, 1965).
To d e t e r m i n e t h e m i n e r a l s (potassium, calcium, magnesium, sulfur, boron, copper, iron, manganese, and zinc) and contaminants (arsenic, cadmium, chromium, and lead), 0.5 g of each sample were digested in nitricperchloric acid (7 mL). After digestion, the samples were diluted to 25 mL using deionized water and stored in plastic tubes at room temperature. Mineral composition and contaminants were analyzed by inductively coupled plasma optical emission spectrometry (ICP-AES, Varian ® ). The nitrogen analyses were carried out by Kjedhal method in hot sulfuric acid extract (Nogueira et al., 2005).

Statistical analysis
For the total soluble solid variables, pH, moisture, fibers, phenolic and mineral compounds, ANOVA was applied separately for each harvest considering a completely randomized design with five treatments and four replications. For the longitudinal variable (yield), a structure was considered to distribute treatments in subdivided plots, in which the plots have the treatments and in the subplots, the time (days after planting).
For the instrumental color variables (L, a*, b*), mass loss, and turgor pressure, ANOVA was performed considering a completely randomized design in a subdivided plot scheme, with five treatments (A, B, C, D, and E) and in the subplots three times (0, 5 and 9 days), with six repetitions.
The verifications of the technique assumptions were carried out, and, when necessary, the Scott-Knott means comparison test was applied for the treatments. All analyses were performed using the R Core Team software (R Core Team, 2020), considering a significance level of 5%.

Yield
Regarding the production of fresh mass, the interaction was significant between time and treatment, indicating that these factors are dependent. Evaluating the developments (Table  1; Figure 1), there was a significant increase in the curly kale yield with the organomineral source (B) supply in the four crops analyzed, with the highest averages.
In the first harvest (60 days), all treatments differed (p≤0.05) from each other, in which the control (A) had the second-lowest average fresh mass production, which was higher only than the organic fertilizer E (comprising sugarcane filter cake). At 90 days, organic fertilizer D (sewage sludge) showed the lowest average fresh mass production, and organic fertilizers C and E were similar to the control (A). For the harvest (120 days), the organomineral fertilizer was superior to all the others. At 150 days after planting, the production of the control treatment (A) showed the lowest average, and the three organic fertilizers (C, D, and E) did not differ from each other, and treatment B (organomineral) showed the highest average fresh mass production (Table 1; Figure 1). The increase in the fresh mass yield with the organomineral source is probably due to the greater immediate availability of nutrients in the soil from the mineral fraction of the fertilizer, while the organic fraction may have influenced the improvement of the physical structure of the soil in the long term. A similar effect was discussed by Mumbach et al. (2020).
In all the analyzed crops, the organomineral fertilizer (B) yield was between 12 and 34% higher than the control treatment (A). The differences between the organomineral source (B) and the organic sources (C, D, and E) were more significant in the first harvest (60 days), with differences between 8 and 57%. At 150 days, despite the tendency to decrease productivity in all treatments, the organomineral source yield remained between 9 and 13% higher than the organic sources (C, D, and E). This positive response to the supply of organic fertilizers in kale was also observed by Balcãu et al. (2012) and Haile & Ayalew (2018).

Microbiological analysis
Microbiological analysis results showed that the kale from all treatments in the three harvests met the standard established by Brazil (2019) and were absent for Escherichia coli. There was also no contamination by total coliforms or thermotolerant coliforms or by Salmonella spp. from any fertilization treatment; that is, the results are under the legislation that establishes the absence of Salmonella spp. in 25 g of sample.
Analyzing 65 samples of certified organic vegetables, Maffei et al. (2013) did not find Salmonella spp. in the samples. However, they found thermotolerant coliforms in 41.5% of organic vegetables. Machado et al. (2006) also did not find Salmonella spp. in vegetables grown under different organic fertilizers.

Physicochemical analysis
There was no significant interaction between the evaluation and treatment days for the variables of mass loss, turgor pressure, and instrumental color (L, a*, and b *), indicating that these factors are independent. Thus, the factors were assessed separately (Table 2; Figure 2), and there was no influence on the type of fertilizer on the parameters evaluated.
There was a significant effect during storage, with paraboloid behavior for the variables mass loss, turgor pressure, and color a* and b*, increasing mass loss ( Figure 2a) and b* (Figure 2e). For the instrumental color L parameter ( Figure  2c), there was a linear behavior, with a 1.09 increase in lightness each day, regardless of the treatment.
Between the fourth and fifth days, a maximum point for turgor (350 Kpa) was observed for kale regardless of the treatment applied. Readings close to zero refer to withered leaves, while the most extensive readings are from leaves not subject to dehydration . Evaluating turgor in kale, Ferreira & Calbo (2010) found 372 Kpa, higher than the values observed in this study, which ranged from 329.41 Kpa to 339.37 Kpa for kale grown using organic fertilizers. The turgor of vegetables is a parameter that indicates the plant quality, in the case of leafy plants, that is, the consumer does not well accept leafy plants with signs of wilting.
Mass loss was obtained for each unit of days, as shown in the second-degree equation in Figure 2a. Studying kale stored in plastic packaging at 10 o C, Simões (2004) reported a 15% mass loss on the 2 nd day of storage.   with signs of yellowing and loss of turgor, causing a consequent short postharvest period (Sanches et al., 2016). According to Novo et al. (2011), aspects of appearance such as size, shape, brightness, and leaf color are the main quality attributes observed by the consumers when buying leafy vegetables. The leaf color attribute is of fundamental importance, as the consumer decides to buy the product based mainly on the appearance, associating it with a freshness indicator without considering the texture, nutritional value, and flavor. Vegetables with greener and brighter leaves are recommended for consumption by professionals in nutrition and gastronomy (Henz & Mattos, 2008).

DAP A (control) B (organomineral) C (organic) D (organic) E (organic)
The analysis of total soluble solids, pH, moisture, fibers, ash, and phenolic compounds of curly kale is shown in Table 3. The average levels of total soluble solids varied in the three harvests but showed no significant

DAP* Treatments CV(%) A (control) B (organomineral) C (organic) D (organic) E (organic)
TSS (  vegetables are estimated to range between 20 and 50% (Finger & França, 2011). Due to the high metabolism of kale, its leaves are highly perishable, Post-harvest losses of vegetables may be related to physiological factors, mechanical damage, or parasites. The accumulated post-harvest losses on difference between treatments. Sanches et al. (2016) found a value of 7.8°Brix in collard leaves. The authors reported that production factors, such as genetic material used, soil type, climate, and cultural treatments, might affect this parameter.
The average pH of the treatments was in the range of 5.62-6.47 and showed significant differences between some treatments, however, according to Menezes et al. (2005), the ideal pH range for vegetables is between 5.0 and 7.0. Sanches et al. (2016) obtained a pH of 5.20 in collard. Concerning kale moisture, the levels varied between 85.33 to 86.27%, with no statistical difference between treatments, which were equivalent to the values already described by Lorenz & Maynard (1988) of 85% for curly kale. Kale presented between 3.34 to 4.34% fiber, with no difference between treatments. Curly kale leaves had a  higher fiber content than collard, which had 1.3 g (Luengo et al., 2011) and 2.0 g/100 g (Lorenz & Maynard, 1988).
The phenolic compounds' content varied between 586.97 and 888.65 mg EAG per 100 g, with no statistical difference between treatments. There was an increase in values for all treatments from the first to the second harvest. The values obtained for curly kale were higher than those found by Rigueira et al. (2016), who obtained phenolic compounds between 173 to 244 mg per 100 g in leaves of another variety of collard. These differences occur since the comparison is made with different varieties. In addition to the fact that collard has a higher moisture content, a dilution of the values of solids and phenolic compounds, can probably occur.
The ash content in kale varied between harvests; higher values were found in the 2 nd and 3 rd harvest. Compared with the collard with 1.3% (Taco, 2011), the curly kale's ash contents are higher.
Analyzing nutrients in kale leaves, the averages of treatments in the three harvest seasons did not show statistically significant differences (p≥0.05) for any analyzed minerals (Table 4). The levels of N, B, Cu, and Fe within the range of values are considered adequate by Trani et al. (2015) for kale, in all treatments used.
The N levels in the three harvests are within the range considered adequate by Trani et al. (2015). The values of protein content ranged from 25.2 to 31.6 g kg -1 , calculated from the total N content (Table 4). For comparison, Thavarajah et al. (2016) obtained values between 16 and 59 g kg -1 by studying kale cultivars; and Thavarajah et al. (2019) found values between 13 and 60 g kg -1 in kale grown in an organic system.
In all the averages, the P levels were higher than the limits considered adequate by Trani et al. (2015) in the first harvest (60 days). The levels in all treatments of the second and third harvests were in the range considered adequate and were equivalent to the values observed by Luengo et al. (2018).
The K contents of the treatments with the different fertilizers are within the appropriate range. However, the control treatment showed values above those considered adequate for K in the first harvest (60 days). In the other harvests (90 and 120 days), the values were within the appropriate ranges.
In the first harvest, the Ca levels were below the appropriate range (Trani et al., 2015) in the control treatments, fertilized with organomineral (B) and organic (C and D) fertilizers. Only treatment with organic fertilizer E was in the range Table 4. Average macronutrient analysis in the kale leaves grown with organomineral and organic fertilizers compared to Trani et al. (2015). Araras, UFSCar, 2019. considered adequate. However, from the second and third harvests (90 and 120 days), the secondary macronutrient's values were in the appropriate range.

Paramters DAP Treatments CV(%) Trani et al. (2015) A (control) B (organomineral) C (organic) D (organic) E (organic)
These adequate values are in agreement with Luengo et al. (2011), who observed calcium levels in kale of 25.0 g kg -1 and Brussels sprouts of 18.2 g kg -1 . The average leaf levels of Mg were below the limits of the appropriate range for all treatments studied in the harvest at 60 days, but were in the appropriate range in the harvest at 90 days and decreased again to values below adequate in the last harvest (120 days). Luengo et al. (2018) also observed leaf levels of Mg below the critical level in 61% of the properties analyzed.
For sulfur, there are no reference values in Trani et al. (2015). However, it is an essential nutrient for brassicas, as it participates in the composition of the cysteine and methionine amino acids and presents high extraction (Cantarella & Montezano, 2010). The nitrogen:sulfur ratio varied from 4.2:1 to 5.3:1 for the treatments evaluated in the first harvest (60 days) and, 3.9:1 to 7.8:1 in the second harvest (90 days), and 5.6:1 to 7.5:1 in the third harvest (120 days). These ratios are below the values generally found in plant tissues, between 12-15:1 (Cantarella & Montezano, 2010). With a higher amount of sulfur in relation to nitrogen, these relationships indicate that the sulfur present in the soil was sufficient to meet the nutritional requirements of kale.
The decreasing order of the macronutrient contents in kale leaves was: N>K>Ca>S>P>Mg, in which nitrogen is the most extracted nutrient, followed by potassium and calcium. This result differs from previous studies conducted with brassicas, in which greater potassium extractions were observed in kale leaves (Aquino et al., 2009;Correa et al., 2013). However, these results are in agreement with Takeishi et al. (2009) for leaf, stem, and cauliflower inflorescence. The sulfur contents were higher than those of phosphorus, and these were higher than those of magnesium. In studies carried out by Aquino et al. (2009), Takeishi et al. (2009), and Correa et al. (2013, sulfur contents were higher than phosphorus and magnesium. In the first harvest, the B values in the leaves were within the appropriate range (Table 5), and from the second and third harvests (90 and 120 days), the values of this micronutrient were below the appropriate range. Luengo et al. (2018) also observed that most of the kale samples from properties in the Federal District were below the appropriate range proposed by Trani Table 5. Average analysis of micronutrients and contaminants in kale leaves grown with organomineral and organic fertilizers compared to Trani et al. (2015). Araras, UFSCar, 2019.

Parameters DAP Treatments CV(%) Trani et al. (2015) A (control) B (organomineral) C (organic) D (organic) E (organic)
Averages followed by distinct letters in the lines differ (p≤0.05) by the Scott-Knott test. * 60, 90, and 120 days. A: control, without fertilizer application; B: organomineral fertilization, class B: organic compound of materials for reuse and recycling of waste generated by human activities, macro and mineral micronutrients; C: organic fertilization, class A: comprising poultry litter, bone meal, industrial slaughterhouse waste, and food and other industrial waste; D: organic fertilization, class B: composing sewage sludge; E: organic fertilization, class A: composing sugar cane filter cake.
et al. (2015). Probably N reduced B concentration in the tissues as there is an antagonism in the absorption of B by the increase in the applied N (Mengel & Kirkby, 2001;Rodrigues et al., 2009). The Cu and Fe levels in the three evaluation periods and all treatments and control (Table 5) were within the appropriate ranges proposed by Trani et al. (2015). Mn average leaf contents were below the limits of the appropriate range for all treatments studied in all evaluation periods. These levels below the adequate one were also observed by Luengo et al. (2018) in samples of different properties. The reduction in Mn absorption can be explained by the presence of N, preferably in the form of nitrate in well-drained soils such as the experimental area. The absorption of this anion leads to an increase in rhizosphere's pH, which can result in decreased absorption of metallic cations such as Mn and Zn (Mengel & Kirkby, 2001).
Zn content was below the adequate level for the control treatment in the first harvest (60 days) and in the appropriate range on the other treatments. However, after harvest at 90 and 120 days, all Zn levels were below the range considered adequate. These results indicate a negative interaction between the absorption of P and Zn, as the excess of the absorption of the macronutrient (Table 5) inhibited the absorption of the micronutrient as already demonstrated by Mengel & Kirkby (2001).
Regarding micronutrients, the decreasing order of leaf contents was Fe>B>Zn>Mn>Cu. These results are in agreement with the contents presented by Luengo et al. (2018).
There is a growing interest in green leaves that contain higher concentrations of Zn, and that can provide a substantial source of this element (Broadley et al., 2010). Zn plays an essential role in blood clotting and protecting DNA from modification, decreasing the risk of cancer (Messias et al., 2015). Brasil (2005) recommends that the daily amount consumed by a healthy adult individual is between 9 to 59 mg of Fe and 7 mg of Zn. Therefore, 100 g of fresh kale would correspond to 35% of the minimum amount of Fe and 7% of the recommended value of Zn.
The contaminating heavy metals (arsenic, cadmium, chromium, and lead) were not detected in curly kale leaves grown using different fertilizers. This indicates that the sources used had no contaminants that could harm the final quality of the vegetable.
The organomineral fertilization promoted more outstanding production of curly kale leaves, regardless of the harvest season. Organic and organomineral fertilizers made it possible to produce curly kale with adequate physicochemical composition, free from microbiological contamination and heavy metals.