Apparent digestibility coefficient of chickpea , maize , high-quality protein maize , and beans diets in juvenile and adult Nile tilapia ( Oreochromis niloticus )

The objective of our study was to assess the apparent digestibility of plant ingredients in diets for juvenile (50 g) and adult (220 g) Nile tilapia (Oreochromis niloticus). Dietary dry matter and protein apparent digestibility coefficients of four plant-derived feedstuffs (chickpea, maize, high-quality maize protein, and beans) were tested. The beans diet had the lowest apparent digestibility coefficient of dry matter (ADCDM) (69.41%), while no significant differences were detected in ADCDM among the other diets; ADCDM was significantly higher in adults compared with juveniles (77.02 vs. 73.76%). Apparent dry matter digestibility coefficient of ingredients (ADCI) was significantly higher in the chickpea (70.48%) and high-quality protein maize (71.09%) ingredients, and lower in the beans (52.79%) ingredient. Apparent dry matter digestibility coefficient of ingredients was significantly higher in juveniles compared with adults (72.56 vs. 56.80%). The protein digestibility of diet (ADCCP) was significantly higher in the reference diet (93.68%), while the lowest corresponded to the maize (87.86%) and beans (87.29%) diets. Significantly lower apparent digestibility coefficient of protein (ADCICP ) was obtained with the highquality maize protein (59.11%) and maize (49.48%) ingredients, while higher ADCICP was obtained with the chickpea and beans ingredients (71.31 and 63.89%, respectively). The apparent digestibility coefficient of ingredient crude protein ADCICP was significantly higher in juveniles compared with adults (67.35 vs. 53.46). Digestibility is generally higher in juveniles, and we recommend using chickpea as an ingredient in diets for Nile tilapia.


Introduction
Diets for fish are usually formulated with different animal and plant ingredients.Animal-derived ingredients are mainly used to satisfy protein requirements, particularly of carnivorous species.These ingredients are generally more expensive, scarce, and less available than ingredients of plant origin (Chamberlain, 1995).Various protein sources, such as coffee pulp (Ulloa Rojas and Verreth, 2003), leucaena leaf meal (Wee and Wang, 1987), cottonseed meal (Lee and Dabrowski, 2002), moringa (Richter et al., 2003), and torula yeast (Olvera-Novoa et al., 2002) have been studied as ingredients in diets for tilapia.
According to FAO Stat (2015), with 1000 Mt, corn is one of the cereals with highest production worldwide, while legumes production is 42 Mt, with more than a half corresponding to beans and 12 Mt to chickpea.Corn, chickpea, and beans are highly available low-cost products in the international market.In his review on the use of plant protein sources in fish diets, Hardy (2010) stressed that, despite the rapid increases in prices of plant meals, the cost per unit protein for plant protein sources remained lower than that of fishmeal protein.
The most important characteristic of any feedstuff is bioavailability, particularly digestible protein, available amino acids, and digestible energy.Understanding the digestibility of ingredients is a basic requirement for diet formulation (Cho and Kaushik, 1990).Coefficients of apparent digestibility provide estimates of nutrient availability in feedstuffs and are used to select ingredients that optimize nutritional value and cost of formulated diets (Fagbenro, 1999).Among the factors known to affect nutrient digestibility in fish are the species, feed size, frequency of feeding, and fish size/age (De Silva and Perera, 1983;Henken et al., 1985;Usmani and Jafri, 2002).He et al. (2013) found that digestibility decreased as the size of Nile tilapia increased.Our objective was to assess the effect of size of Nile tilapia (juvenile and adult) on the apparent digestibility coefficient of chickpea, maize, highquality protein maize, and beans as ingredients in diets.There are no studies on the digestibility of these ingredients in Nile tilapia.

Material and Methods
The feeding trial was conducted in Guasave, Sinaloa, México.Four plastic tanks (40 L) per treatment were used and supplied with filtered, sterilized freshwater.The tanks were covered with plastic mesh to prevent escape of fish.Cultivation conditions were: photoperiod of 14 h light:10 h dark, 27±1 °C, and dissolved oxygen >4 mg/L.
Five diets were prepared (Table 1).One was a reference diet (30% crude protein, CP; and 8% lipids, LIP) and four were experimental diets.The experimental diets contained 69.0% of the ingredients of the reference diet, 30% of the tested ingredient, and 1% chromium oxide as a marker.The ingredients were ground with a mill (Laboratory Mill 3610, Perten Instruments, Hägersten, Sweden) and sieved through a 460 μm mesh.Diets were prepared with an extruder and dried at 45 °C until their moisture content was 8-10%.Afterwards, pellets were ground to a size appropriate for the size of fish, and stored at -20 °C until required.Triplicate samples of each diet were used for chemical analysis.The tested ingredients were chickpea (Cicer arietinum) variety Blanco Sinaloa 92, white maize (Zea mays), high-quality maize protein (HQMP), and azufrado (sulphur yellow) beans (Phaseolus vulgaris).The diets were analysed for their proximate composition (Table 1) by standard methods (AOAC, 1995) and for their energy content, using an adiabatic calorimeter (PARR, Moline, Illinois, USA).Juvenile (50±2 g) and adult (220±5 g) tilapia were used.Each experimental tank (0.50 × 0.35 × 0.30 m; total of 40) was stocked with one fish obtained from the CIIDIR-IPN hatchery.Four tanks per diet were randomly assigned.Tilapias were acclimated to the experimental conditions for one week, consuming the reference diet to apparent satiation, which was delivered twice daily (08.00 and 16.00 h).After acclimation, fish were fed to apparent satiation twice daily (08.00 and 16.00 h), for 54 days.Faeces were collected from each tank three hours after each feeding, using a Pasteur pipette with a siphon (Jones and De Silva, 1997).Faeces were rinsed with distilled water, lyophilized, and stored at −70 °C until further analysis.Protein and chromic oxide in the faeces and diets were evaluated simultaneously by the method of Bolin et al. (1952), a modified micro-Kjieldahl method (Nieto et al., 1997).
The percentage of apparent dry matter digestibility (ADC DM ) and apparent protein digestibility (ADC CP ) were calculated using the equations of Maynard et al. (1979), cited by Cruz-Suarez et al. (2001): × [(% CP in faeces)/(% Cr 2 O 3 in faeces)], in which CP is the crude protein content.The percentage of apparent dry matter and protein digestibility of ingredients

Results
There were significant differences in ADC DM among diets and among juveniles and adults.No significant interaction between the main factors was detected.Apparent digestibility coefficients of dry matter varied from 71.32% (beans diet) to 74.86% (chickpea diet) in juveniles, and from 67.5% (beans diet) to 85.37% (reference diet) in adults.A significant lower value of ADC DM was detected in the beans diet, while no significant differences were detected among the other diets (Figure 1a).Apparent digestibility coefficient of dry matter was significantly higher in adults (Figure 1b).There were significant differences in ADC I among diets and among juveniles and adults.No significant interaction between the main factors was detected.Apparent dry matter digestibility coefficients of ingredients varied from 60.92% (beans diet) to 76.33% (chickpea diet) in juveniles, and from 44.67% (beans diet) to 64.64% (chickpea diet) in adults.Values of ADC I were significantly higher in the chickpea and maize QPM diets, and lower in the beans diet (Figure 2a).Apparent dry matter digestibility coefficient of   ingredients was significantly higher in juveniles (Figure 2b).There were significant differences in ADC CP among diets, but no significant difference was detected between juveniles and adults.No significant interaction between the main factors was detected.Apparent digestibility coefficients of crude protein varied from 86.94% (beans diet) to 92.73% (reference diet) in juveniles and from 87.42% (maize diet) to 94.63% (reference diet) in adults.Significantly higher ADC CP was obtained with the reference diet, while the lowest ADC CP was obtained with the maize and beans diets (Figure 3).There were significant differences in ADC ICP among diets and among juveniles and adults.
No significant interaction between the main factors was detected.Apparent crude protein digestibility coefficients of ingredients varied from 60.42% (maize diet) to 75.25% (chickpea diet) in juveniles and from 38.54% (maize diet) to 67.37% (chickpea diet) in adults.Significantly lower ADC ICP was obtained with the maize QPM and maize diets, while a higher ADC ICP was obtained with the chickpea and beans diets (Figure 4a).Apparent crude protein digestibility coefficient of ingredients was significantly higher in juveniles (Figure 4b).

Discussion
Apparent digestibility of dry matter and protein in the tested ingredients depended on the type of ingredient and the size of the Nile tilapia.The differences in ADC of ingredients may be explained by differences in chemical composition, which in turn is determined by the origin or processing of the feed ingredients (Köprücü and 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 favourably to those obtained by studies with other freshwater tropical fish species.Nile tilapia has a relatively long gastrointestinal tract and faecal collection by natural voidance is the only realistic option for determining the apparent digestibility coefficient (Suresh, 2003;Sklan et al., 2004).Ramos et al. (2012) showed that low digestibility of cotton and cocoa bran in tilapia might be a consequence of the presence of large amounts of fibre and anti-nutritional factors in these by-products.In our study, even though beans contain high levels of protein, the ADC I was low in juveniles and adults.Aparicio-Fernández et al. (2005) reported that beans contained 4.1 mg g -1 of tannins after thermal processing.The adverse effect of tannic acid on tilapia was studied by Al-Owafeir (1999), who found a significant reduction in the growth of tilapia that were fed diets containing as low as 0.27% tannic acid.However, Becker and Makkar (1998) reported that 2% quebracho tannins (condensed tannins) was tolerated without reducing the growth of common carp, whereas similar levels of hydrolysable tannins (tannic acid) reduced acceptability of the feed after four weeks.
The presence on tannins, phytic acid, and trypsin inhibitors may be the cause of low digestibility of beans  ( Gaber, 2006).While chickpeas belong to the same family as beans, in our study, chickpeas produced the highest ADC values.According to El-Adawy (2002), cooking (heat treatments) improved the in vitro protein digestibility and protein efficiency ratio of chickpeas.The improvement in digestibility results from denaturation of protein, destruction of the trypsin inhibitor, or reduction of tannins and phytic acid (Liener, 1980;El-Adawy, 2002;Gaber, 2006).The extrusion process resulted in higher protein productive value of chickpea meal for gilthead bream Sparus aurata (Adamidou et al., 2011).In our study, heating during the extrusion process probably resulted in higher digestibility of chickpeas, as an ingredient.
Several investigations have examined the potential of maize gluten meal.The ADC DM of maize gluten meal reported by Köprücü et al. (2004), Köprücü and Özdemir (2005), and Guimarães et al. (2008) for Nile tilapia (88.0-91.0%)was higher than the our results (74-77%).Köprücü and Özdemir (2005) suggested that the differences in ADC of nutrients and energy come from the differences in chemical composition, origin, and processing of the feed ingredients.
In our study, ADC DM , ADC I , and ADC ICP indicated higher digestibility in juveniles.Similarly, He et al. (2013) found that efficiencies of digestible protein, energy, and amino acids significantly decreased as tilapia grew larger.Alvarez-González et al. (2008) indicated that the quality and quantity of digestive enzymes of fish vary with age.

Conclusions
The digestibility of the tested ingredients is generally higher in juvenile Nile tilapia.In particular, considering that the price of chickpea is lower (US $950/t) than that of fishmeal (US$ 2388.6), we recommend using chickpea as an ingredient in their diets.
was determined by the method described byCho and Slinger (1979), as follows: %ADC I = [(100 × %ADC DM of TD) -((100 -% TI) × %ADC DM of RD)]/(%TI) %ADC ICP = [(100 × %ADC CP of TD × %CP in TD) -((100 -%TI) × %ADC CP of RD × %CP in RD)]/(%TI × %CP in TI), in which ADC I is the apparent dry matter digestibility of the ingredient; ADC DM of TD is the apparent dry matter digestibility of the tested diet; TI is the tested ingredient; ADC DM of RD is the apparent dry matter digestibility of the reference diet; ADC ICP is the apparent protein digestibility of the ingredient; ADC CP of TD is the apparent protein digestibility of the tested diet; CP in TD is the concentration of protein in the tested diet; TI is the tested ingredient; ADC CP of RD is the apparent protein digestibility of the reference diet; CP in RD is the concentration of protein of the reference diet; and CP in TI is the concentration of protein in the tested ingredient.A two-factor, completely randomized experimental design, with four replicates per treatment, was used.Values of digestibility were tested for normality and variance homogeneity.A two-way ANOVA and a Tukey's multiplerange test were used to compare means values of digestibility of diets and ingredients among juveniles and adults.Diets and the size of the fish were the main factors in the two-way ANOVA.Statistica 7.0 software (StatSoft, Tulsa) was used for the tests, with significance set at P<0.05.
Results of the two-way ANOVA are included in the figure.Means followed by different letters are significantly different.

Figure 1 -
Figure 1 -Effect of diets (A) and tilapia size (B) on apparent digestibility coefficient of dry matter (%ADC DM ).

Figure 2 -
Figure 2 -Effect of diets (A) and tilapia size (B) on apparent digestibility coefficient of ingredients (%ADC I ).

Figure 4 -
Figure 4 -Effect of diets (A) and tilapia size (B) on apparent digestibility coefficient of ingredients protein (%ADC ICP ).

Figure 3 -
Figure 3 -Effect of diets on apparent digestibility coefficient of dietary protein (%ADC CP ).

Table 1 -
Formulation and proximal composition of experimental diets for O. niloticus