Flavored Amazonic pirarucu (Arapaima giga) waste flour (salted and sweet) for inclusion in food products

Abstract

This study aimed to develop flavored flours (salty and sweet) from Amazonic pirarucu waste (Arapaima gigas), include them in extruded snacks, and evaluate the nutritional, physicochemical, microbiological, and acceptance characteristics of these products. A standard flour was elaborated with pirarucu carcass, which presented 54.42% of protein and 7.24% of lipids, and from this, flavored flours were elaborated (salty and sweet). The standard flour had higher levels of protein, calcium, and phosphorus; and the salted one had higher levels of lipids. The fatty acids present in greater quantities were oleic (average of 32.21%), linolenic (average of 20.74%), and palmitic (average of 17.81%). The flavored flours were better accepted than the standard flour, for all sensory attributes and purchase intention. The snacks with sweet flour, despite better results in the sensory attributes of color, aroma, and flavor, were the ones that presented the lowest content of protein and ash, when compared to those with inclusion of standard flour. It is concluded that the pirarucu waste can be used for producing flavored flours and extruded snacks, with the purpose of improved food products.

Introduction

In recent decades, the consumption of fish protein has been increasing, and in 2018 the per capita consumption reached 20.5 kg. This is due to population growth and the growing concern for healthy food with high nutritional quality (FAO 2018). The pirarucu (Arapaima gigas) is a species native to the Amazon basin (Amazonas-Brazil), and is the largest freshwater fish, reaching an average of 3 m in length and can reach up to 200 kg body weight in the natural environment (Hernández et al. 2017). The pirarucu is a very appreciated species in the market due to the potential that the species presents for pisciculture, for having a large size and high market value; however, as its fishing is restricted in Brazil due to the ecological control, there is a stimulus to the production of this fish in captivity (Martins et al. 2017).

The food industry is constantly searching for innovative foods that not only please the palate but are also healthy. Fish flour is rich in polyunsaturated fatty acids of the omega 3 and omega 6 families, which, which prevent cardiovascular diseases, such as arteriosclerosis and cancer, besides acting on the immune system and on anti-inflammatory processes (Singh et al. 2014). These nutrients can be included in human nutrition in an indirect way through the inclusion of the flours, besides the biological value of the protein, with several essential amino acids in desirable amounts to meet daily requirements (WHO 2007).

Currently, the byproducts of fish processing are mainly used through the manufacture of flour for addition to animal feed, representing 68%, and the remainder has unknown or inadequate destination (Souza et al. 2021). Fish flour comes as an alternative to improve the use of byproducts from the processing line of the beneficiation industry, because with the increase in fish production, the amount of waste generated is also high, and this raw material has high nutritional quality and can be used for human consumption. Fish flour is produced by cooking, pressing, grinding, drying, and crushing the raw material (Souza et al. 2017). There are several technologies for producing the flours, such as by smoking the carcasses, however, cooking is the most common method (Justen et al. 2017). Currently, these flours have been studied for human food (Souza et al. 2017). Some authors have evaluated the use of fish flours in extruded snacks (Goes et al. 2015) and lasagnas (Kimura et al. 2017).

Thus, the production of fish flour is an alternative for a better use of the raw material. Therefore, this work aimed to elaborate a flavored pirarucu carcass flour (sweet and salty), as well as its inclusion in extruded snacks, evaluating the nutritional, microbiological, and sensory characteristics of the flours and snacks.

Material and methods

The pirarucu carcasses (spine or backbone with the remaining filleting meat) from cultivation were purchased in Macapá-AP-Brazil (0°2′4''N51°3′60''W), packed, frozen (-18 °C) and transported by air to the Fish Technology Laboratory at Fazenda Experimental de Iguatemi (FEI), belonging to the State University of Maringá (UEM- Maringá-PR-Brazil).

Preparation of flavored flours

The carcasses were thawed at 7 °C for 12 h and then used in the preparation of the flours. First, according to Souza et al. (2017), a standard flour was developed using the pirarucu filleting waste that used 0.5% BHT (butylated hydroxyl toluene) as an antioxidant. From the standard flour without flavor, two types of flavored flours were prepared, one salted and one sweet, following the methodologies described below: for the salted flour was used 53.26% standard pirarucu meal, 8.78% sesame, 1.59% sesame oil, 0.34% monosodium glutamate, 0.79% salt, 2.66% nori, 2.00% fermented soy sauce, 0.26% oregano, 0.26% dehydrated garlic, 5.32% dehydrated onion, 5.32% dried tomato, 0.79% parsley, 5.32% chimichurri, and 13.31% flaxseed. The pirarucu flour was mixed with sesame oil and fermented soy sauce and baked (90 °C) until a homogeneous mass was obtained, and then the dehydrated ingredients were added and baked for another 10 min. For the preparation of the sweet flour, 37.97% pirarucu flour, 37.97% brown sugar, 5.10% cinnamon, 12.65% cocoa powder, 0.25% cloves, 3.79% grated ginger, and 2.27% vanilla were used. First, the pirarucu meal and brown sugar were mixed in 100 mL of water and cooked (90 °C) until a homogeneous mass was obtained. Then the other ingredients were added to the process and cooked for another 10 min.

The obtained doughs were dehydrated in an oven (model brand) at 60 °C for 24 h, and then ground in a knife-type grinder (Willye—model TE-650). After finishing the elaboration of the flavored flours, they were packed and frozen (-18 °C) until the moment of the analyses.

Elaboration of extruded snacks from pirarucu flours

Four types of extruded corn snacks were prepared, being one control, without pirarucu flour (no inclusion) and the others with the inclusion of the flours with pirarucu carcass (standard, salted, and sweet) (Fig. 1). A 10% of fish flour (standard, salted, and sweet) was added to 90% corn grits, and these mixtures were individually added directly to the inlet of the extrusion system using a small sample feeder. For extrusion, the following process parameters were used: Die Plate: two holes with central cut; main motor current: 15–22 A; cutting knife inverter: five A; and homogenizer rotameter: 20 mL/min, and extrusion temperature of 110 ºC (Justen et al. 2017). The snacks were extruded in an Inbramaq extruder (Model IB-50). After finishing the preparation of the snacks, all analyzes were performed immediately.

Fig. 1.

Extruded snacks with the inclusion of pirarucu carcass flours. A Without inclusion; B Standard; C Salted; D Sweet

Proximate composition

The analysis of moisture, ash, and protein was performed according to the AOAC (2005) methodology. For lipid analysis the methodology described by Bligh and Dyer (1959) was followed.

Calcium and phosphorus

Calcium and phosphorus analyses were performed by acid digestion and quantification by flame atomic absorption spectrometry (Zhou et al. 1998).

pH and water activity (Aw)

To the pH analysis, 10 g samples were homogenized in 90 mL of distilled water and determined in a pH meter (DM 22, Digimed). The determination of the water activity (Aw) was performed using the Labswift- Novasina apparatus.

Instrumental color

Colorimeter MINOLTA CR-10 (Minolta Camera Co, Osaka, Japan) was used for determining the instrumental color, where L* defines the brightness (L* = 0 black and L* = 100 white), a* (red-green component), and b* (yellow-blue component).

Amino acid profile

Characterization of the amino acid profile was performed according to the methodology described by White et al. (1986).

Fatty acid profile

The lipids were extracted according to the method of Bligh and Dyer (1959), and transmethylated according to the method of Hartman and Lago (1973). The fatty acid esters were isolated and analyzed using a Shimadzu gas chromatograph, model GC-FID 2010, using an RT-x Wax Polyethylene glycol column (30 m length × 0.25 mm ID × 0.25 µm film thickness). The carrier gas was Helium (He), and the injection flow rate was 1 mL / min divided into 1:20. The initial temperature for the flame column was set at 80 °C, held for five minutes, then raised to 190 °C at a rate of 5 °C/min, held for five minutes, then raised to 220 °C at a rate of 2 °C/min, held for five minutes, and finally raised to 240 °C at a rate of 5 °C/min, held for five minutes. The temperature of the injector and detector used was 250 ºC. Fatty acid identification was performed by comparing the retention times of the methyl esters in the samples with those of the authentic standards (FAMEs). The relative composition of fatty acid methyl esters (FAMEs) was quantified, and the data were presented as weight percent for FAME composition.

From the fatty acid profile, the nutritional quality of the lipid fraction of pirarucu flours was determined through the atherogenicity index (AI), thrombogenicity index (TI), and the ratio between hypocholesterolemic and hypercholesterolemic fatty acids (HH) (Ulbricht and Southgate 1991). AI, TI, and HH were calculated using the following equations: (C12:0 + 4 × C14:0 + C16:0)/ MUFA + n − 6 + n − 3), (C14:0 + C16:0 + C18:0)/[(0.5 × MUFA) + (0.5 × n6) + (3 × n − 3) + (n − 3/n − 6)], and (C18:1n − 9 + C18:2n − 6 + C20:4n-6 + C18:3n − 3 + C20:5n − 3 + C22:5n − 3 + C22:6n6)/(C14:0 + C16:0), respectively.

Microbiological analysis

The microbiological analyses for coliforms at 35 and 45 °C, Salmonella sp. and coagulase positive Staphylococcus aureus were performed according to Franco (2008). The results were expressed as most probable number (MPN) for coliforms at 35 and 45 °C, absence, or presence for Salmonella sp, and in colony forming units per gram (CFU/g) for coagulase positive Staphylococcus aureus.

Sensory analysis of flavored flours and extruded snacks

An affective acceptance test was carried out with 100 untrained consumers and regular consumers of fish products, of both genders, aged 18–50 years, who received guidance on the sensory analysis procedure.

Samples of the three flours (standard, salted, and sweet) and the four types of snacks, without inclusion of flours (control), and snacks with inclusion of flours (standard, salted, and sweet) were prepared at different times and submitted to sensory analysis separately. The consumers received the samples in plastic cups, randomly coded and a cup with water to rinse the mouth between each sample. A 9-point hedonic scale was used, with the extremes: 1 (I disliked it very much) and 9 (I liked it very much) (Dutcosky 2011). The purchase intention was also evaluated using a 5-point scale, where 5 represented the maximum score "would certainly buy" and 1 represented the minimum score "would certainly not buy. This research was approved by the Human Ethics Committee (Copep) of the State University of Maringá (Paraná, Brazil) under protocol No. 458.151/2013.

Statistical analysis

The experiments were repeated three times on three different days and all analyses were performed in triplicate. The results were submitted to analysis of variance (ANOVA) and the means were compared using Tukey's test, considering 5% probability using software SAS (Inst. Inc. Cary, NC, USA), version 9 (2010).

Results and discussion

Flavored flours

There was a significant difference (p < 0.05%) for the moisture values of the pirarucu carcass flours (Table 1), and the salted flour was different from the standard flour. The salted flour showed lower moisture values (3.35%) than the standard flour (4.23%) and the sweet flour (4.07%). This may be associated with the inclusion of the different ingredients of each type of flour produced.

Table 1.

Proximate composition, mineral, pH, and Aw of flavored flours

Composition	Flours			P value	CV (%)
	Standard	Salted	Sweet
Moisture (%)	4.23 ± 0.18a	3.35 ± 0.22b	4.07 ± 0.15a	0.0020	4.56
Proteins (%)	50.46 ± 6.90a	26.63 ± 4.63b	23.71 ± 6.18b	< 0.0001	3.29
Lipid (%)	7.80 ± 0.87b	9.07 ± 1.11a	3.47 ± 1.37c	< 0.0001	7.00
Ash (%)	35.23 ± 4.69a	20.44 ± 2.98b	15.85 ± 3.83c	< 0.0001	1.09
Calcium (%)	5.44 ± 1.09a	1.47 ± 0.76b	1.33 ± 0.79b	< 0.0001	2.25
Phosphorus (%)	7.02 ± 1.38a	2.28 ± 0.92b	1.61 ± 1.05b	< 0.0001	1.94
pH	7.07 ± 0.19b	6.60 ± 0.27c	7.93 ± 0.31a	< 0.0001	1.57
Aw	0.22 ± 0.02b	0.33 ± 0.03a	0.25 ± 0.02b	0.0347	14.39

Means ± standard deviation in the same line followed by different lowercase letters indicate significant difference (p < 0.05) by Tukey´s test. CV = Coefficient of Variation

The protein content of the standard flour was higher than the other samples (p < 0.05%). The standard flour had 50.46% of protein while the salted flour had 26.63% and the sweet one, 23.71%. As for the lipid content, the salted flour had 9.07%, and was significantly higher (p < 0.05%) than the sweet (3.47%) and the standard (7.80%). This is probably due to the amount of sesame oil and flaxseed added that provided the increase in lipid content.

Souza et al. (2017), evaluated the proximate composition of flours produced using the carcass of tilapia and salmon, as well as other species. The moisture values ranged from 1.78 to 4.86%, protein values ranged from 44.63 to 83.28%, and lipid contents ranged from 3.98 to 18.81% and according to the same authors the differences are due to the use of different fish species, each with its own chemical characteristics. Oliveira et al. (2015) developed flour using the mechanically separated meat of Brazilian catfish (Brachyplatystoma vaillantii) and found a moisture content of 12%, 76.16% protein, 7.72% lipids, and 3.95% ash. These values are different from those found in this study due to the different processes used in the manufacture of the flours and due to the differences in the raw material, because they are distinct fish species.

The standard flour presented higher contents of calcium and phosphorus in relation to the salty and sweet ones (p < 0.05%). This occurred due to the amount of skeletal bone present in the standard flour, providing higher mineral contents.

There was a significant difference in ash content (p < 0.05%) between the flavored and standard flours. As the mineral contents (calcium and phosphorus) were significantly higher in the standard flour, because it is composed only of cooked bone, it showed higher ash content (35.23%) compared to salted (20.44%) and sweet (15.85%) flour (Table 1).

Senapati et al. (2016) showed 73% protein, 6.10% moisture, and 2.8% lipid content in fatty fish (Sardinella longiceps) meal without deboning to increase mineral content such as calcium 3.26% and phosphorus 4.14%. On the other hand, Tiwari et al. (2021) observed in fishmeal from indigenous species, average calcium values of 1.85% and phosphorus values of 0.79% for Puntius sophore and 2.09% calcium and 1.01% phosphorus values for Amblypharyngodon mola. It is worth mentioning that in our study, only the pirarucu carcasses were used and not the muscles, which is why there is a difference in the mineral values found between the studies.

The pH of the flours showed significant difference (p < 0.05%) with the standard flour showed average pH values of 7.07, the salty flour average was of 6.60, and the sweet flour average was 7.93 (Table 1). As each type of flour is composed of different ingredients, they may have influenced the differences in pH. Foods with pH above 6.0 are classified as foods of low acidity. In this case, greater care should be taken, due to the possibility of growth of pathogenic bacteria, generating greater care during product storage (Franco and Landgraf 2008). However, the flour with the highest pH was the sweet-flavored, requiring greater care regarding storage time, although the water activity index should also be considered. In this flour, the Aw was significantly lower (0.25) than in the salty-flavored (0.33). The low Aw indicates that the flours have greater stability.

In the standard flour, 19 amino acids were identified, of which ten are essential (Table 2). The essential amino acids identified were arginine (3.26%), histidine (0.89%), phenylalanine (1.84%), isoleucine (1.90%), leucine (3.02%), lysine (3.40%), methionine (1.32%), threonine (2.00%), tryptophan (0.31%), and valine (2.03%). Furuya et al. (2004) stated that lysine (5–7%) is one of the most abundant essential amino acids in fish; however, glutamic acid and glycine were in the highest proportions in fish bone flours.

Table 2.

Amino acid profile of the standard flour

Amino acids	%
Aspartic acid	2.49 ± 0.10
Glutamic acid	5.52 ± 0.31
Alanine	3.35 ± 0.26
Arginine*	3.26 ± 0.09
Cysteine	0.65 ± 0.22
Phenylalanine*	1.84 ± 0.16
Glycine	5.61 ± 0.21
Histidine*	0.89 ± 0.18
Isoleucine*	1.90 ± 0.26
Leucine*	3.02 ± 0.21
Lysine*	3.40 ± 0.16
Methionine*	1.32 ± 0.30
Proline	3.53 ± 0.15
Serine	1.74 ± 0.28
Taurine	0.11 ± 0.11
Tyrosine	1.45 ± 0.26
Threonine*	2.00 ± 0.23
Tryptophan*	0.31 ± 0.17
Valine*	2.03 ± 0.11
Total	44.42

^*Essential amino acids

When evaluating flours made from processing by-products of different fish species, Souza et al. (2017) reported that in tilapia and salmon flours the amino acids found in higher proportions, above 0.075% of protein, were lysine, glutamic acid, aspartic acid, glycine, leucine, and phenylalanine + tyrosine. These amino acids are also found in pirarucu flour, but in lower percentages.

The pirarucu flour, as to its amino acid profile, meets the requirements for adults, according to the WHO (2007). The flour has five amino acids (isoleucine 1.90%, leucine 3.02%, lysine 3.4%, threonine 2.0%, and valine 2.03%) within the values needed to meet the requirements of an adult, and three amino acids with levels close to those required (histidine 0.89%, methionine 1.32%, and tryptophan 0.31%). The requirement to supply the amount of essential amino acids for an adult is 1.6% histidine, 1.3% isoleucine, 1.9% leucine, 1.6% lysine, 1.7% methionine + cystine, 0.9% threonine, 0.5% tryptophan and 1.3% valine (WHO 2007). Therefore, the standard flour has a good nutritional quality.

Table 3 shows the fatty acid profile of the pirarucu flours. The main fatty acids present in the samples were palmitic, stearic, oleic, and linoleic. These fatty acids were also predominant in the study of Souza et al. (2017) in tilapia bone meal. Concerning to saturated fatty acids, palmitic acid was the most abundant (average of 17.81%), followed by stearic acid (average of 12.35%), and the sweet flour had the highest content of these fatty acids (21.56 and 19.50%, respectively). The main monounsaturated fatty acid was oleic (average of 32.21%) and the polyunsaturated fatty acid was linoleic acid, and the salted flour had the highest content (28.20%). Nutritious monounsaturated fats can maintain normal heart rhythm and reduce the risk of certain types of cancer. Consumption of monounsaturated fats helps regulate insulin and blood sugar levels, which is beneficial to the individual, especially for people with diabetes. This effect is due to its good content of phytosterols, substances that help in the elimination of cholesterol and also in the reduction of the abdominal circumference (Damasceno et al. 2013).

Table 3.

Fatty acid profile of standard, salted, and sweet pirarucu flours

Fatty acid (%)	Flours
	Standard	Salted	Sweet
Lauric/ C12:0	0.11 ± 0.23a	0.04 ± 0.04b	0.10 ± 0.09a
Myristic/ C14:0	1.28 ± 0.18a	0.46 ± 0.11c	0.83 ± 0.21b
Pentadecyl/ C15:0	0.28 ± 0.14a	0.09 ± 0.22b	0.20 ± 0.17a
Palmitic/ C16:0	19.47 ± 0.27a	12.42 ± 0.31b	21.56 ± 0.30a
Margaric/ C17:0	0.50 ± 0.32a	0.19 ± 0.27b	0.40 ± 0.12a
Stearic/ C18:0	10.39 ± 0.11b	7.18 ± 0.19c	19.50 ± 0.19a
Arachidic/ C20:0	0.34 ± 0.08b	0.50 ± 0.26a	0.66 ± 0.26a
Behenic/ C22:0	0.13 ± 0.10b	0.20 ± 0.08a	0.16 ± 0.18a
Lignoceric/ C24:0	0.11 ± 0.22a	0.15 ± 0.25a	0.13 ± 0.05a
Palmitoleic/ C16:1 ω7	3.42 ± 0.19a	1.05 ± 0.08a	2.20 ± 0.24a
Cis-10-heptadecenoic/ C17:1	0.27 ± 0.26a	0.11 ± 0.11a	0.18 ± 0.20a
Oleic/ C18:1 ω9	32.36 ± 0.14a	30.92 ± 0.10a	33.36 ± 0.28a
Vacenic/ C18:1 ω7	2.15 ± 0.13a	1.30 ± 0.26a	1.26 ± 0.19a
Gondoic/ C20:1 ω9	0.49 ± 0.04a	0.31 ± 0.22a	0.35 ± 0.16a
Erucic/ C22:1 ω9	–	0.09 ± 0.17a	–
Alpha Linolenic/ C18:3 ω3	1.22 ± 0.09a	14.81 ± 0.05a	0.86 ± 0.14a
Stearidonic/ C18:4 ω3	0.08 ± 0.26a	–	0.04 ± 0.17a
di-homo-alpha-linolenic/ C20:3 ω3	0.22 ± 0.24a	0.06 ± 0.19a	0.13 ± 0.19a
Eicosapentaenoic EPA/ C20:5 ω3	0.17 ± 0.14a	0.09 ± 0.25a	0.08 ± 0.23a
Linoleic/ C18:2 ω6	19.88 ± 0.09a	28.20 ± 0.10a	14.14 ± 0.31a
Gamma Linolenic GLA/ C18:3 ω6	0.16 ± 0.12a	0.05 ± 0.05a	0.11 ± 0.07a
Conjugated Linoleic CLA/ C18:2 ω6	0.97 ± 0.22a	0.27 ± 0.18a	0.59 ± 0.27a
Eicosadienoic/ C20:2 ω6	0.37 ± 0.18a	0.12 ± 0.19a	0.23 ± 0.06a
di-homo-gamma-Linolenic DGLA/ C20:3 ω6	0.81 ± 0.27a	0,23 ± 0.21a	0.47 ± 0.05a
Arachidic/ C20:4 ω6	1.51 ± 0.23a	0.41 ± 0.17a	0.90 ± 0.15a
Others	3.30	0.76	1.54
Saturated fatty acids (SFAs)	32.61	21.23	43.54
Monounsaturated fatty acids (MUFAs)	38.69	33.78	37.35
Polyunsaturated fatty acids (PUFAs)	25.39	44.24	17.55
ω3	1.69	14.96	1.11
ω6	23.70	29.28	16.44
ω9	32.85	31.32	33.71
ω6/ω3	14.02	1.96	14.81
PUFAs/SFAs	1.96	3.53	1.26
AI	0.36	0.18	0.45
TI	0.85	0.26	1.38
HH	1.72	3.59	1.58

Means ± standard deviation in the same line followed by different lowercase letters indicate significant difference (p < 0.05) by Tukey´s test

In their study on the inclusion of different percentages (0, 5, 10, and 15%) of tilapia meal in roll, Chambó et al. (2018) evaluated the profile of fatty acids with an average of 25% palmitic acid, 15% stearic acid, 42% oleic acid, and 3% linoleic acid. The values are different from those presented by the flours of the present study possibly because other types of fish were used and also because they are different products.

It can be seen in Table 3 that the standard flour had high levels of omega 9, which is a monounsaturated fatty acid. Omega 9 is a monounsaturated fatty acid, and is related to healthier triglyceride levels, and also helps to decrease the levels of total blood cholesterol, LDL (bad cholesterol), and also increase the HDL (good cholesterol) (Delgado et al. 2017).

The ratio of fatty acids n-6 / n-3 is used as a criterion to assess the quality of the fat, which should be below 4 (FAO 2010). In this study, the salted flour obtained values of 1.96 and is in accordance with the criterion; however, the other formulations had an average of 14.41, values higher than 4. It is recommended in the ideal human diet that the proportion of polyunsaturated fatty acids (PUFA)/saturated fatty acids (SFA) is at least 0.45 (Wood et al. 2004). In this study, all formulations were within the recommended proportion and the salted flour presented the highest ratio of 3.53. It is likely that the best results presented by the salted flour may be due to the addition of sesame oil and flaxseed in its formulation.

The indices of atherogenicity (AI) and thrombogenicity (TI) and the ratios of hypocholesterolemic to hypercholesterolemic (HH) fatty acids are also used to evaluate the quality of the lipid fraction. According to Turan et al. (2007), in relation to the indices of atherogenicity (AI) and thrombogenicity (TI), the lower the indices, the higher the quality of the anti-atherogenic fatty acids present in lipids. On the other hand, the higher the ratio between hypocholesterolemic and hypercholesterolemic fatty acids (HH index), the better the nutritional quality of the fatty acids present in the product. Thus, the best option taking into consideration these nutritional parameters, would be to use the salted flour that presented an AI index of 0.18, TI of 0.26, and HH of 3.59, which would be more indicated for the regulation of coronary heart disease.

Through the microbiological analysis of the standard and flavored flour it was observed that they were produced in the appropriate conditions of hygiene, being suitable for human consumption. The values of coliforms at 35 and 45 °C were lower than 3 NMP/g, the count of coagulase positive Staphylococcus was lower than 1 × 10² CFU/g and the analysis of Salmonella spp. was absent in all samples.

Figure 2a shows the results of the sensory analysis of the standard and flavored flours, and it shows that the flavored samples received higher scores for all attributes (p < 0.05%). For the attributes of color, aroma, texture, and flavor there was no significant difference (p > 0.05%) for the salted (7.21 to 7.80) and sweet (6.98 to 7.58) flours, but the scores were significantly higher than those of the standard flour (4.22 to 5.71). It can be seen that the scores assigned by the consumers corresponded to I liked moderately and I liked very much.

Fig. 2.

Sensory analysis of pirarucu flours (A), extruded snack (B)

Likewise, the purchase intention of the flavored flours was higher in relation to the standard one. The scores attributed by the consumers for the salty and sweet flavored flours were 3.9 and 3.87, respectively, indicating a probable purchase of the products if they were available in the stores.

Extruded snacks

The inclusion of fish flour in extruded snacks provided a reduction in moisture and increase in protein, lipids, ashes, calcium, and phosphorus contents (p < 0.05%), as shown in Table 4.

Table 4.

Proximate composition, mineral, pH, Aw, and color parameters of extruded snacks

Composition	Flours				P value
	Without inclusion	Standard	Salted	Sweet		CV (%)
Moisture (%)	8.80 ± 0.42a	7.63 ± 0.23b	6.87 ± 0.31c	6.94 ± 0.29c	< 0.0001	2.49
Proteins (%)	7.78 ± 0.60c	10.97 ± 0.56a	9.61 ± 0.36ab	9.41 ± 0.35b	0.0007	5.78
Lipid (%)	4.82 ± 0.28b	6.03 ± 0.24a	6.13 ± 0.26a	5.11 ± 0.22ab	0.0099	7.44
Ash (%)	2.53 ± 0.23ab	2.95 ± 0.27ab	3.01 ± 0.28a	1.40 ± 0.39b	0.0397	4.64
Calcium (%)	0.008 ± 0.05c	0.040 ± 0.06a	0.025 ± 0.03ab	0.015 ± 0.03bc	0.0010	5.33
Phosphorus (%)	0.001 ± 0.0d	0.020 ± 0.01a	0.009 ± 0.04b	0.004 ± 0.02bc	0.0042	4.96
pH	6.27 ± 0.19c	7.15 ± 0.19a	6.86 ± 0.10b	7.13 ± 0.13a	< 0.0001	0.55
Aw	0.58 ± 0.002a	0.47 ± 0.03b	0.47 ± 0.03b	0.47 ± 0.03b	0.0052	0.14
L*	79.37 ± 1.15b	84.78 ± 2.02a	78.04 ± 1.18b	73.85 ± 1.87c	< 0.0001	1.13
a*	5.60 ± 0.85a	2.90 ± 0.73c	5.06 ± 1.19a	3.77 ± 0.62b	< 0.0001	7.25
b*	34.70 ± 2.48a	29.05 ± 1.56b	27.92 ± 1.52b	20.03 ± 2.74c	< 0.0001	3.48

Means ± standard deviation in the same line followed by different lowercase letters indicate significant difference (p < 0.05) by Tukey´s test. CV = Coefficient of Variation

The moisture of the extruded snacks with the inclusion of pirarucu flours varied between 8.80% for the snacks without fishmeal inclusion and 6.87% for the snacks with salted fishmeal inclusion. The sweet and salted snacks presented the lowest moisture contents, which is interesting because lower moisture contents are important in extruded snacks, as it gives the product more crunchiness (Souza et al. 2021).

The inclusion of standard flour in the extruded snacks resulted in a 41% increase in protein content (p < 0.05%), based on the snacks without fish flour inclusion. With the use of flavored flours, salted and sweet, the increment in protein was lower for the sweet (20.95%) and for the salted, which did not differ from the one with inclusion of standard flour (23.52%). Thus, with the addition of the flours, the extruded snacks presented protein enrichment.

When evaluating snacks with inclusion of flour of different fish species, Goes et al. (2015) reported that the control (no inclusion) showed 6.82% protein, while the samples with addition of 9% flour of different fish species showed higher protein contents (11.85% with added tuna, 10.21% with added sardine, 9.80% with added tilapia, and 9.21% with added salmon). The authors reported that this difference is due to the type of raw material and fish species used in flour preparation.

For lipid contents there was no significant difference (p > 0.05%) between the formulation without fishmeal and the sweet one; however, the extrudates with added standard and salted flour were different from the formulation without flour (p < 0.05%). The values ranged from 4.82 to 6.13.

The snacks with added sweet flour showed lower ash content, despite differing only from the salted-flour snacks. It can be inferred that the ingredients added in the preparation of the sweet flour diluted the percentage of minerals present, while in the salted flour the salt and glutamate present in the formulation may have provided the increase in ash content.

In the work of Justen et al. (2017), evaluating the inclusion of different levels of tilapia (Oreochromis niloticus) flavored meal in extruded snacks, there was a reduction of up to 4.8% of moisture, with an increase of up to 40.22% of protein, 39.71% of lipids, and 135.29% of ash due to the inclusion levels (0% to 12%) of fish meal in the snacks.

Monteiro et al. (2018) used tilapia flour to replace wheat flour in bread and showed an increase in protein, lipids, and ash values. Patimah et al. (2019) showed 11.89 protein values, 21.30% lipids, and 45.19% carbohydrates in cookies with flying fish flour.

Regarding the analyzed minerals, fish flour addition in snacks significantly increased the calcium and phosphorus contents, with the standard flour being higher (0.040% and 0.020%, respectively) than the others, differing only in the calcium level of snacks with sweet flour (Table 4). In the salted flour, ingredients with high calcium contents, such as nori seaweed and sesame, were added.

Snacks with 3 to 12% tilapia flour inclusion showed a linear increase from 0.15 to 0.73% of calcium and from 0.13 to 0.44% of phosphorus (Justen et al. 2017). This behavior was also observed in this study where calcium increased from 0.008% for the snacks without added fishmeal, to an average value of 0.026% in the snacks with fishmeal. Phosphorus showed the same behavior with an increase from 0.001% to on average 0.011% in the snacks without added and with added flour, respectively.

There was a significant difference (p < 0.05%) between pH values for snacks that averaged between 6.27 and 7.15. Fishmeal inclusion resulted in lower Aw values in the snacks (p < 0.05%), as observed in Table 4, and there was a correlation between moisture and Aw values, since the samples without fishmeal inclusion showed the highest values of moisture and Aw, respectively. Similar results were reported by Justen et al. (2017), where snacks with added fishmeal showed average Aw values of 0.52 while in the samples without added fishmeal the Aw values averaged 0.57. According to these authors, fishmeal inclusion in extruded snacks provides a reduction in Aw, thus ensuring better microbiological stability of the product. The lipid profile in extruded snacks and the low water activity may promote lipid oxidation, which entail the need for lightproof oxygen-free packaging during storage (Oliveira et al. 2015), or the addition of additives and antioxidants (Justen et al. 2017).

Regarding the instrumental color, the snacks with standard flour showed higher brightness (Table 4) compared to the others (p < 0.05%). The sweet-flour snacks showed lower brightness (73.85), possibly due to the inclusion of brown sugar, chocolate, and cinnamon powder added in the formulation that caused a darker coloration in the snacks. Justen et al. (2017) reported L* values of 79.2 in snacks without fishmeal, values close to those of the present study (79.37). When analyzing the extruded snacks with tilapia flour inclusion, the same authors reported that a reduction in brightness occurs with increasing levels of flour, making the snacks darker (L* 79.2% with 0% and L* 73.63% with 12%). Thus, the type of raw material used and the flour elaboration methodology, as well as the inclusion levels used in the extruded snacks interfere in their luminosity.

The red intensity (a* coordinate) was significantly higher in the snacks without flour inclusion (5.60) and with salted flour (5.06), and the snacks with standard flour presented significantly lower a* value (2.90).

The snacks with sweet flour inclusion (20.03) showed lower yellow values (b* coordinate), while the snacks without flour inclusion had more intense yellow coloration (34.70) possibly due to the higher amount of corn grits that have yellow coloration (Fig. 1). On the other hand, snacks without the inclusion of fish flour showed lower b* values (28.8), compared to snacks prepared with 3% (31.15) to 12% (29.03) of fish flour inclusion in the work developed by Justen et al. (2017).

The microbiological analysis of the extruded snacks showed a probable number (MPN) of coliforms at 35 and 45 °C lower than 3 MPN/g for all treatments; for the count of coagulase positive Staphylococcus values were lower than 1 × 10² CFU/g and absence of Salmonella spp, showing that the production occurred under adequate conditions of hygiene and sanitation, and were suitable for human consumption.

Figure 2b shows the results for the sensory analysis of the snacks. The attributes color and aroma were similar for the snacks without inclusion of flour and with sweet flour (p > 0.05%), whose scores were significantly higher than the other treatments. For the texture attribute, the snacks without added fish flour and the sweet snacks showed significant differences (p < 0.05%) in relation to the standard and salted snacks. Regarding flavor, only the snacks with salted flour showed significant difference (p < 0.05%) compared to snacks without added flour. The scores were lower when salted flour was added, not differing from snacks with added standard flour (p < 0.05%).

Patimah et al. (2019) obtained good acceptance when they added 15% flying fish flour in cookies. Similarly, Monteiro et al. (2018) performed sensory analysis of bread with the addition of tilapia-waste flour and the authors suggested that the replacement of 10% of wheat flour did not affect the overall taste of the products, enabling the nutritional enrichment of bread with good sensory acceptance by consumers. Above this percentage, the tasters did not accept the products very well because, according to the authors, the inclusion of high levels of fish sources in foods is problematic, due to the fishy taste and odor generated mainly by free fatty acids and volatile sulfur.

In analyzing extruded snacks of Indian major carp (Labeo rohita) fish, Singh et al. (2014) showed that the highest organoleptic score was acquired with the addition of approximately 20% fish meal.

The snacks without added flour and with sweet flour presented the highest scores for purchase intention (3.47 and 3.33, respectively), and the production of sweet, extruded snacks is an alternative to the use of pirarucu by-product.

Conclusion

It is concluded that the Amazonic pirarucu waste can be used for the production of flavored flours, for improved food products, adding value to the productive chain of the pirarucu. The flavored flours can be produced from the standard flour, presenting excellent nutritional, microbiological, and sensory quality.

The inclusion of pirarucu flour in extruded snacks provides an increase in protein, lipids, calcium, and phosphorus content. The snacks with sweet flour presented better sensory acceptance.

These results demonstrate that the industries can invest in new ways to maximize the use of the pirarucu waste for its application in food products.

Author contributions

S.S.Correa: Formal analysis, Writing—original draft. G.G.O.: Formal analysis. F.V.d.S.: Formal analysis. M.F.C.: Formal analysis. L.F.d.S.A.: Formal analysis. M.A.M.: Formal analysis. E.G.: Supervision. Writing—review & editing. E.S.d.R.G.: Writing—review & editing. A.C.F.: Supervision. Writing—review & editing. M.L.R.d.S.: Project administration, Resources, Supervision, Writing—review & editing.

Availability of data and material

All data generated or analysed during this study are included in this published article.

Declaration

Conflict of interest

The authors declare no conflict of interest.

Ethics approval

'Ethical approval for this study was obtained from the Ethics committee of the State University of Maringá under number CAAE: No. 458.151/2013.