Effect of fermented, hardened, and dehulled of chickpea (Cicer arietinum) meals in digestibility and antinutrients in diets for tilapia (Oreochromis niloticus)

Among the most typical feed sources for tilapia, plants represent a low-cost source in substituting for traditional high-cost feed ingredients. Fermentation, hardening and dehulling are common grains processing techniques to make plant nutrients available and more digestible to fish. Apparent digestibility coefficients (ADC) of dry matter and protein, and antinutrients (phytic acid and tannins) in fermented, hardened and dehulled chickpea (Cicer arietinum) meals were determined for juvenile Nile tilapia (Oreochromis niloticus). The highest ADC was obtained with processed (fermented, hardened and dehulled) chickpea meals compared with non-processed. Results indicated that fermentation increased the protein content by 13.1%, decreased the content of ash and phytic acid (47.5 and 45%, respectively), and increased the ingredient apparent digestibility of dry matter (ADM) by 23.2%, and the ingredient apparent digestibility of protein (ADP) by 41.9%. Dehulling meal increased the protein (5.7%) and lipid (6.4%) content of chickpea grains; decreased fiber, ash and tannin content (75.3%, 19.1%, and 84.5%, respectively); and increased ADM by 12.8%, and ADP by 10.4%. We conclude that fermented, hardened and dehulled chickpea meals represent a potential alternative in diets for juvenile O. niloticus. Additional key words: aquafeeds; plant-based feed ingredients; bioprocessing; antinutritional compounds; tilapia. Abbreviations used: ADC (apparent digestibility coefficients); ADM (apparent digestibility of dry matter); ADP (apparent digestibility of protein); SSF (solid-state fermentation). Authors’ contributions: Conceived and designed the experiments: RGD and HRG. Performed the experiments: FJVG, RGD and BLCR. Analyzed the data: FJVG, MGU and HRG. All authors wrote and approved the final manuscript. Citation: Valdez-González, F. J.; Gutiérrez-Dorado, R.; García-Ulloa, M.; Cuevas-Rodríguez, B. L.; Rodríguez-González, H. (2018). Effect of fermented, hardened, and dehulled of chickpea (Cicer arietinum) meals in digestibility and antinutrients in diets for tilapia (Oreochromis niloticus). Spanish Journal of Agricultural Research, Volume 16, Issue 1, e0605. https://doi.org/10.5424/ sjar/2018161-11830 Received: 03 Jun 2017. Accepted: 22 Mar 2018 Copyright © 2018 INIA. This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 International (CC-by 4.0) License. Funding: Instituto Politécnico Nacional (SIP20170366; SIP20160482; COFAA and EDI); Consejo Nacional de Ciencia y Tecnología (Grant: 266171). FJVG is a recipient of a PhD fellowship from CONACYT. Competing interests: The authors have declared that no competing interests exist. Correspondence should be addressed to Hervey Rodríguez González: hrodriguezg@ipn.mx; hervey23@hotmail.com


Introduction
Since feed represents up to 70% of production costs in tilapia cultivation, a priority area of research is substituting traditional high cost ingredients such as fish meal by the incorporation of low-cost agro-industry by-products (Tacon & Metian, 2008;Montoya-Mejía et al., 2016). Typical feed sources for tilapia include fish and poultry meals (Abdel-Warith et al., 2001;Gonzales et al., 2007), soybean meal (Azaza et al., 2009) and other plant sources (El-Saidy & Gaber, 2003;Guimarães et al., 2008). To more accurately formulate diets, the nutritional value of new ingredients should be validated to make them more available to fish. In the case of vegetable by-products, different processing techniques are applied on seeds or grains to promote better nutrient and energy digestibility (Valdez-González et al., 2016), and to reduce or avoid the presence of antinutritional components (Valdez-González et al., 2013). Fermentation , dehulling (Booth et al., 2001), and hardening (Reyes-Moreno et al., 2000), are the most representing processing techniques commonly used on plant-based ingredients. Then, an important prerequisite to validate the nutritional value of new ingredients is the determination of apparent digestibility coefficients (ADC) (Davies et al., 2011) of dry matter (ADM) and protein (ADP), among others, with different bioprocessing techniques.
The biotechnological processing of ingredients by solid-state fermentation (SSF) represents a technological alternative for processing a great variety of legumes and/or cereals to improve their nutritional and nutraceutical values (Angulo-Bejarano et al., 2008). In general, SSF can be performed with Rhizopus sp. fungi since an important function of the fungus during fermentation is the synthesis of enzymes, which can hydrolyze some of the substrate constituents and contribute to the development of a desirable texture, flavor and aroma of the product Sánchez-Magaña et al., 2014).
SSF is a low-cost alternative processing technique for preparing aquaculture diets for aquatic species. It yields high quality products with a minimal degradation of nutrients, and has a significant improvement in digestibility and biological value of proteins Guzmán-Uriarte et al., 2013). It has also been proved that fermentation induce favorable changes in the legumes antinutrients, such as reduction in the activity of enzymatic inhibitors (i.e. phytates and tannins; Reyes-Moreno et al., 2004).
On the other hand, a factor limiting the use of legumes as ingredients in diets is hardening, which occurs when legumes are stored under adverse conditions of high temperature and relative humidity (Reyes-Moreno et al., 2000;Medina-Godoy et al., 2011). Longer periods of cooking and lower nutritional value characterize legumes with this deficiency Medina-Godoy et al., 2011). Hardening affects a high percentage of grain during storage, causing hardened grain being considered a by-product that is sold at low prices. Another limiting factor for using legumes in diets is their content of antinutrients (Guillaume et al., 2004;Glencross et al., 2004Glencross et al., , 2007. The effect of vegetal antinutritional factors has been less studied in fish than in higher vertebrates (Guillaume et al., 2004;Valdez-González et al., 2013).
To overcome those limitations, processes need being implemented for, on one hand, improvement and utilization of large quantities of legumes and cereals stored for a long time (Drew et al., 2007;Adamidou et al., 2011;Valdez-González et al., 2013) and, on the other hand, reducing the content of fiber, tannins and increase protein digestibility by for, example, the elimination of the hull or dehulling . The objective of the present study was to test the effect of the processes of fermentation, hardening and dehulling on the chemical composition of chickpea (Cicer arietinum) grains, in vivo digestibility of ingredients, and antinutrients (phytic acid and tannins) in diets for Nile tilapia Oreochromis niloticus including flours of processed and non-processed grain chickpea. There are no antecedents in the literature of investigations like the one conducted in this study.

Preparation of chickpea flours
Fresh and hardened chickpea variety ˈBlanco Sinaloa 92ˈ, were used. Hardening of chickpea was applied according the procedure of Reyes-Moreno & Paredes-López (1993) under laboratory conditions, with slight modifications. The hardening of seeds was produced using accelerated storage (37 ± 1 °C, relative humidity of 100%, 15 d). Flours of fresh and hardened chickpea were prepared utilizing SSF. Accordingly, whole grains were soaked for 16 hours in a 0.4% glacial acetic acid solution (pH = 3.1), drained and manually dehulled. The hull grains were added at the end of the fermentation process and drying of the samples. Cotyledons were cooked in distilled water at 90 °C for 30 min and then cooled at 25 °C for 4 h. The substrate was inoculated with a suspension of Rhizopus oligosporus NRRL2710 (1 × 10 6 spores/mL) in 15 × 25 cm polyethylene bags with small holes (3 cm separation). Lots of bags with cotyledons were incubated (Riossa, mod EC-33, Mexico) for fermentation at 34.9 ºC during 51 h. Subsequently, samples were dried in an oven with forced air circulation (50 ºC, 24 h). Finally, samples were milled (Tecator Mill, mod 1083, Sweden) to obtain a flour mesh #80 (0.180 mm).
The fresh and hardened non-fermented chickpea flours were prepared by milling the grains in a ½ HP electric mill until approximately four fragments per chickpea were obtained. After this, the hulls were removed using an electric fan and the fragments were milled to obtain a flour mesh #80 (0.180 mm). Dehulled chickpea was fermented separately, and finally, milled with the hulls.

Preparation of diets
A reference diet and eight experimental diets were prepared. In each experimental diet, 30% of the ingredients of the reference diet were replaced by the corresponding chickpea flour ( Table 1). The experimental diets were: Fresh whole chickpea, fresh hulled chickpea, hardened whole chickpea, hardened hulled chickpea, fresh whole fermented chickpea, fresh hulled fermented chickpea, hardened whole fermented chickpea, hardened hulled fermented chickpea.
The ingredients were ground through a mesh #40 (0.425 mm), and subsequently mixed and homogenized. 3 Chromium oxide (1%) was added as an inert marker to determine feed digestibility. Feed was prepared in a meat mill Torrey® Mexico (Monterrey, Mexico).

Digestibility assays
Nile tilapias were confined in 27 rectangular plastic tanks (270 L each) using a stocking density of 6 fish (25.0 ± 2.6 g) per experimental tank. Every experimental unit received continuous aeration, keeping dissolved oxygen at 7.25 ± 0.7 mg/L and water temperature at 26 ± 2 °C. Diets were tested with three replicates. Feed was offered to apparent satiation twice a day (08:00 and 15:00 h). Two hours after each feeding, feces were collected with a plastic siphon, washed with distilled water, and placed at -40 °C. One gram of feces (dry weight) were recollected during 30 days. Subsequently, feces were lyophilized and analyzed to determine the content of chromium oxide and proteins.

Chemical analysis
Chemical analysis of the ingredients, diets and feces were performed according to standard methods by AOAC (1995). MicroKjeldahl method was used to determine protein, and determination of nitrogen was conducted in a Kjeltec system (Mod 1009 and 1002, Tecator, Sweden). For determination of lipids, extraction with petroleum ether in a Soxtec system (Mod 1043, Tecator, Sweden) was utilized. Fiber was determined by drying and burning of the sample after extraction using 0.5 M H 2 SO 4 and 0.5 M NaOH. Ash content was determined by calcination of the sample in a Muffle furnace (Thermolyne 6000) at 600 °C for 5 h, and the energy content was determined by an adiabatic calorimeter (Table 2). Chromic oxide in the feces and diets were evaluated by the method of Bolin et al. (1952) and using the equation proposed by Furukawa & Tsukuhara (1966).

Phytic acid
Phytic acid was determined following the procedure of Latta & Eskin (1980). The extraction was performed by shaking (400 rpm at 25 °C during 1 h) 1 g of flour, adding 20 mL of HCl at 2.4%. After this, the suspension was centrifuged (20,000 × g at 25 °C for 5 min) and the supernatant was kept in a freezer. Subsequently, a glass column (0.7 × 27 cm) packed with glass fiber and 0.5 g of ion exchange resin (Bio-Rad) was used. The column was washed with 15 mL 5% HCl and then with 20 ml of deionized water. The supernatant was diluted 1:25 and 10 mL were added in the column. Once the fluid went through the column, 15 mL of 0.1 M NaCl were added and the eluate was discarded. A 25 mL vessel was placed under the column and 15 mL 0.7 M NaCl were added to collect the eluate. After this, deionized water was added to complete a volume of 25 mL. Three milliliters were taken from this solution, and 3 mL of deionized water + 1 mL reagent Wade (0.15 g FeCl 3 .6H 2 O + 1.5 g of sulfosalicylic acid in 500 mL deionized water) were added, shaking thoroughly. The tubes were centrifuged (5000 × g at 25°C for 10 min) and the supernatant was separated; following this, color  (100); menadione (5); thiamine (60); riboflavin (25); pyridoxine HCl (50); pantothenic acid (75); niacin (40); biotin (1); inositol (400); cyanocobalamin (0.2); folic acid (10) Table 3 shows that fermentation increased protein (13.1%) and ash (47.5%) contents of chickpea meal. Dehulling increased the protein (5.7%) and lipid (6.4%) contents; and decreased fiber and ash contents (75.3 and 19.1%, respectively) in chickpea grains.
The effect of fermentation and dehulling on ADM were significant (p<0.001) ( Table 5). There were no significant interactions among the three factors. Average values of ADM in non-fermented and fermented chickpea diets were 58.5 ± 6.2% and 72.1 ± 2.9%, respectively, improving ADM by 23.2% by was measured in a spectrophotometer (Spectronic 21D mod, Milton Roy, USA) at 500 nm.

Tannins
The content of tannin was determined by the method of vanillin proposed by Price et al. (1978) with modifications. Extraction was carried out within 24 h after milling using approximately 1 g of sample and 10 mL of a 1% HCl solution in methanol. The suspension was kept on shaking for 40 min at room temperature and centrifuged (20,000 × g, 30 ºC, 20 min). Five milliliters of reagent of vanillin (50:50 v/v 1% vanillin in methanol and 8% HCl in methanol) were added to 1 mL of supernatant at a rate of 1 mL/ min. After this, the suspension was kept in the dark for 20 min and read in a spectrophotometer (Spectronic 21 mod D Milton Roy, USA) at 500 nm. A blank solution, zero absorbance, was prepared with 1 mL methanol by adding 5 mL of 4% HCl at a rate of 1 mL/min. A standard curve of catechin was plotted and the results were reported as equivalents of catechin.

Statistical analysis
Values of digestibility were tested for normality and variance homogeneity. A multifactorial ANOVA and

Discussion
The increase of protein in chickpea during the fermentation process is related to the protein synthesis caused by proliferation and increase in biomass of R. oligosporus (Paredes-López et al., 1991;Reyes-Moreno et al., 2000;Cuevas-Rodríguez et al., 2004). In this sense, Sánchez-Magaña et al. (2014) reported that the increase in protein content is associated with the decrease of other constituents, which might have been lost by leaching during the initial steps of fermentation or might have been consumed by the fungus for its own growth. There are similar reports when using R. oligosporus with various substrates, such as fresh and hardened chickpea (Reyes-Moreno et al., 2000;Angulo-Bejarano et al., 2008).
The decrease of lipid content in fermented chickpea can be explained as a consequence of the oxidation and fermenting. Mean values of ADM for diets using whole and dehulled chickpea were respectively 61.3 ± 9.5% and 69.2 ± 6.2%, indicating an increase of 12.8% associated with dehulling.
Fermentation and dehulling effects on ADP were significant (p<0.001) ( Table 5). No significant interactions among the three factors were detected. Values of ADP in non-fermented and fermented chickpea grains averaged 65.1 ± 5.9% and 92.4 ± 3.1%, respectively, showing an increment of 41.9% caused by fermentation. Mean values of ADP for whole and dehulled chickpea were 74.9 ± 17.1% and 82.7 ± 14.5% respectively, increasing 10.4% by dehulling. The combination of fermentation and dehulling in grains of chickpea caused an increase of 40.27% for the ADM, and 58.20% for the ADP with respect to non-fermented and not dehulled treatments.  utilization of fatty acids as main source of energy by the fungus (Ruiz-Terán & Owens, 1996). There are similar reports when using R. oligosporus and various substrates: mix corn/soybean (Mugula & Lyimo, 2000); soybean (Ruíz-Terán & Owens, 1996), quality protein maize  and common bean (Guzmán-Uriarte et al., 2013). Dehulling chickpea decrease the content of ash and fiber mainly as a consequence of hulls containing certain minerals (calcium, phosphorus, magnesium, iron, potassium) and a high concentration of fiber (Laurena et al., 1986;Williams & Singh, 1987). The increase in lipid content in dehulled chickpea may be an effect of concentration, considering the loss of other components such as ash and fiber.
Antinutritional factors are chemical elements contained in vegetables affecting the digestibility and the metabolism of energy sources (proteins and lipids) in artificial diets used for animal nutrition (Allan et al., 1999;Valdez-González et al., 2013). Deshpande & Cheryan (1984) considered that the interaction of phytate with proteins, vitamins and several minerals is one of the factors limiting the nutritional value of vegetable meals. In this study, we found that both fermentation and dehulling significantly decreased the content of phytic acid and tannins in chickpea. The content of phytic acid observed in this study was similar to those reported by Jukanti et al. (2012) for chickpea (1.0 mg/100 g dw of sample). Fermentation decreased phytic acid content most likely as a consequence of an increase in phytase activity (an enzyme that synthesizes R. oligosporus), and the soaking-cooking-leaching process used for fermentation (Laurena et al., 1986). Similar results were reported by Sánchez-Magaña et al. (2014) using R. oligosporus and chickpea as a substrate.
It is known that most of the tannin content is found in the hull of legume grains (Egounlety & Aworh, 2003;Guillaume et al., 2004). Adewusi & Osuntogun (1991) mentioned that dehulling reduces 95% of the tannins content in legumes. In close agreement with those authors, in this investigation the tannin content in chickpea grains was substantially reduced (85%) by removing hulls.
The use of highly digestible feed ingredients for cultivable aquatic species is highly recommended (Silva et al., 2013;. There are studies affirming that fermentation causes an increased availability of the protein content in legumes Angulo-Bejarano et al., 2008;Sánchez-Magaña et al., 2014). Higher digestibility associated with fermentation is a consequence of improved nutritional balance of the ingredients and proteolytic activity of fungus, releasing peptides of the substrate and increasing the susceptibility of the protein to enzymatic activity (Cuevas-Hernández et al., 1999). Our study shows that diets based on fermented-dehulled chickpea were easily digested by juvenile Nile tilapia, mainly as a consequence of the protein quality produced by R. oligosporus, in combination with a reduction and blocking of antinutrients activity.
Apparent digestibility of dry matter and protein in the tested ingredients depended on the type of ingredient. The differences in ADM may be explained by differences in chemical composition, which in turn is determined by the origin and processing of the feed ingredients (Köprücü & Özdemir, 2005). Nile tilapia is apparently able to assimilate a wide variety of feedstuffs (Davies et al., 2011) and digestibility data in this study compare favorably with those obtained by studies with other freshwater tropical fish species. Yigit & Olmez (2011) showed that low digestibility of canola meal in tilapia might be a consequence of the presence of large amounts of fiber and antinutritional factors in these byproducts. In our study, fermentation and dehulling of grain chickpea improved digestibility of dry matter and protein, most likely as a consequence of reduction in the content of phytic acid and tannins. Bairagi et al. (2004) reported that the increase in protein digestibility in fermented diets is related to the reduction of antinutrients, such as tannins and phytic acid. Yuan et al. (2013) reported that fermented soybean flour in diets for juvenile Chinese sucker Myxocyprinus asiaticus showed high digestibility without causing adverse effects on growth and body composition. On the other hand, Ramachandran & Ray (2004) reported that the fermented seed flour of the legume Phaseolus mungo, using the bacterium Bacillus sp. in diets for Labeo rohita, showed values of ADP of 86.76%. In contrast, previous studies have shown that ingredients derived from legumes without any process show low levels of digestibility (Bureau et al., 1999;Lara-Flores et al., 2007;Azaza et al., 2009;Phumee et al., 2011;Collins et al., 2012).
In this study, higher protein digestibility coefficients were obtained in comparison to those reported by Adamidou et al. (2011) for Sparus aurata fed with diets prepared with chickpea (Cicer arietinum L.). They reported ADP of 80% for chickpea flour, while in the present investigation, using dehulled chickpea flour, resulted in ADP as high as 94.5 and 95.6%. Booth et al. (2001) reported ADP of 81.1% in diets for Australian silver perch (Bydanus bydanus) using chickpea, which is lower than that obtained in the present study. Tiril et al. (2009) reported a coefficient of 80.6% digestibility of protein using extruded chickpea in rainbow trout (Oncorhynchus mykiss).
Various studies indicate that grains with high phytate content may decrease protein digestibility (Vielma et al., 7 1998;Guillaume et al., 2004;Chan et al., 2008). Levels higher than 0.5 mg/100 g dw of sample of phytic acid in diets for rainbow trout and salmonids have been shown to decrease protein digestibility (Reddy et al., 1989). In this investigation, fermentation caused a reduction in the content of phytic acid in chickpea and an increase in protein digestibility of diets for Nile tilapia.
We observed that dehulling reduced the content of tannins in grains of chickpea. Several reports indicate that enzymatic inhibition caused by tannins, decrease the digestibility of nitrogenous nutrients (Allan & Rowland, 1994;Booth et al., 2001), thus causing a low protein digestibility (Reichert et al., 1980). Pinto et al. (2000) mentioned that levels of tannins higher than 0.63 mg/g significantly affect the digestibility of dry matter, protein, and lipids in Nile tilapia. In addition, the removal of hull may cause a decrease in endogenous antinutrients of the hull (Booth et al., 2001), so as in structural polysaccharides, which in high concentrations are known to reduce dry matter digestibility in fish diets (Wilson, 1994;McGoogan & Reigh, 1996). There are reports that levels of fiber lower than 3 g/100 g, improve the protein digestibility in Nile tilapia (Dioundick & Stom, 1990;Lanna et al., 2004). In our study, dehulling allowed to reduce the levels of fiber below that level, which could contribute to increase ADP.
We found that fermented and dehulled chickpea (C. arietinum) meals represent a potential alternative feed ingredient for preparation of diets for juvenile Nile tilapia O. niloticus. The combination of fermentation and dehulling increases the digestibility coefficients of dry matter and protein in chickpea meals. Hardening did not affect chemical composition or digestibility of the diets, indicating that low priced hardened chickpea could be used in diets. The results obtained in this study showed that fermentation and the removal of chickpea hull, increased protein content, decreased antinutrients, and favored the digestibility of dry matter and protein ingredients in diets for Nile tilapia.