Japanese quail ( Coturnix coturnix japonica ) is a poultry species raised worldwide for several purposes, mainly meat and egg production ( Dalmau, 1994 ; Ayaşan, 2013 ). In several European countries a hybrid strain is also raised in game farms for release in hun­ting preserves in order to ensure hunting bags, usually through put-and-take shooting ( Puigcerver et al. , 2007 ; Sanchez-Donoso et al. , 2012 ). As a mono­gastric granivorous species, balanced feeds and rations used to feed quails are formulated with inclusion of at least 50% of cereal grains, mainly corn ( Konca & Büyükkiliç, 2013 ; Kasmani & Mehri, 2015 ; Mota et al. , 2015 ). Several cereals have been tested as substitutes for corn in feed formulation for growing and laying quails, or to complement corn-based basal rations ( Konca & Büyükkiliç, 2013 ; Ashour et al. , 2015 ). Among other experiments, the use of triticale ( Güçlü & Işçan, 2003 ; Ragab & Namra, 2010 ; Wahed et al. , 2010 ), barley ( Oğuz et al. , 2011 ; Toprak & Yilmaz, 2012 ; Kianfar et al. , 2013 ), wheat ( Sarica et al. , 2009 ; Yasar & Gok, 2014 ; Mehraei Hamzekolaei et al. , 2016 ), oat ( Yasar & Gok, 2014 ) or rice ( Cardoso et al. , 2011 ) has often been cited.

Cereal grains are rich in carbohydrates and have been employed in animal feeding since antiquity. While bread wheat and especially durum wheat are mainly used for human food, other cereals such as barley and triticale are preferably used to feed lives­tock. About 85% of barley production is dedicated to animal feeding ( Feedipedia, 2016 ). When barley is used in diet formulation for laying quails, crude pro­tein must remain at least at 18% to ensure an adequate egg quality and weight ( Toprak & Yilmaz, 2012 ). Repla­cing corn with fermented barley or wheat improves gro­wing chicks performance up to three weeks of age ( Yasar & Gok, 2014 ). Wheat efficiency in keeping growing performance in quails when it substitutes corn under high ambient temperature has been described ( MacLeod & Dabutha, 1997 ), and ground wheat permits growing quails to balance nutrients intake when it is supplied along with a concentrate feed ( Canoğullari et al. , 2004 ). However, inclusion of wheat in quail rations is subjec­ted to limits because its soluble non-starch polysacchari­des content increases the viscosity of digesta ( Sarica et al. , 2009 ). Triticale is an artificial alloploid made from hybrids between tetraploid wheat and rye, and later chromosome duplication with colchicine to increase flower fertility. Interest of triticale is increasing in farmers seeking a low input or sustainable agriculture ( Palta et al. , 2010 ). It has higher protein content than most cereals and a lysine concentration greater than 30% respect to wheat ( FEDNA, 2016 ). It has been demonstrated that the good nutritional properties of triticale makes it a good alternative to corn in quail grower diets ( Ebrahimi et al. , 2017 ). It has been also reported that 40 to 60% triticale can be used to replace corn in balanced feeds without affecting the performance of laying quails ( Güçlü & Işçan, 2003 ).

Cereal grains are generally ground when used in commercial balanced feeds for quails and these are presented as meal or in pelleted form. Moreover, processing of cereals before inclusion in balanced feed improves its nutritive value and quail performance ( Kianfar et al. , 2013 ; Yasar & Gok, 2014 ). However, the use of whole grain in poultry has recently increa­sed to reduce feed processing and transporting costs ( Cumming, 1994 ; Forbes & Covasa, 1995 ; Svihus, 2001 ; Bennett et al. , 2002 ), and in response to the demands of consumers for more 'natural' feeding systems ( Gabriel et al. , 2008 ). Furthermore, some research point out that feeding whole grains to poultry improves gut health ( Ferket, 2000 ) or even that in free choice feeding experiments, the type or form of the grain do not affect the production performance of birds; however, these may affect intake and the efficiency of utilization of some nutrients ( Forbes & Covasa, 1995 ). There are husbandry practices in which intake and preference of whole cereal grains by quails may be relevant. For instance, in the finishing phase of the game farming of this species to prepare individuals for the transition from the farm to the wild, when birds are provided with variable amounts of whole cereal grains (Dal­mau, 1994). To date, voluntary intake and preference of whole-grain cereals have not been investigated in quails, and only the effect of choosing between ground emmer, oat ( Konca & Büyükkiliç, 2013 ) or wheat ( Canoğullari et al. , 2004 ; Konca & Büyükkiliç, 2013 ) and a balanced feed has been studied. In this context, the aim of this study was to assess voluntary intake and preference of four whole-grain cereals (barley, durum and bread wheat, and triticale) in the Japanese quail, as well as the influence of sex and body weight on both feeding patterns.

The research was conducted at the Research and Teaching Farm of the Higher Technical School of Agricultural Engineering (ETSIA) of the University of Seville, Spain (37º 21' 36.3" N, 5º 56' 23.9" W), during February and March of 2015. Bird management and handling were performed according to the Directive 2010/63/EU on the protection of animals used for scientific purposes ( European Parliament and Coun­cil, 2010 ). Two experiments were performed, one to investigate the intake level of barley ( Hordeum vulga­re L.), durum wheat ( Triticum turgidum subsp. durum (Desf.) Husn.), bread wheat ( Triticum aestivum L. subsp. aestivum ) and triticale (× Triticosecale Witt­mack) whole-grain cereals, and another to assess preference of quails for these cereals.

To our knowledge, literature on whole-grain cereal voluntary intake and preference in quails is scarce. In fact, this is the first work that undertakes this assessment simultaneously for the four studied cereals: durum and bread wheat, triticale and barley.

Regarding the lower intake of whole-grain of barley in both experiments (especially in the preference trial), it should be taken into account that barley is the only hulled cereal of this study. Barley hull consists mainly of cellulose, hemicellulose, and lignin. Therefore crude fibre content is higher, and the digestible energy of barley is lower compared to the remaining cereals (of hulled or naked kernel) in this work. Hull may also act as a diluent of available nutrients or by physically or chemically inhibiting nutrient digestion and absorp­tion ( Sharifi et al. , 2012 ).

Furthermore, when comparing voluntary intake by Japanese quails among feeds including different cereals, Yasar & Gok (2014) observed that feed intake of diets including fermented barley was significantly reduced in comparison to diets including fermented wheat. This agrees with the lower feed intake of barley grain in the present study in comparison to wheat grains (Table 2). Yasar & Gok (2014) explain this effect on the basis of differences in chemical composition and physical texture of both cereal grains that would lead to different nutrient digestibility. In particular, they found that these intake differences were highly correlated with the reduced total dietary fibre, non-starch polysaccharides, and β-glucan contents of the diet that included wheat compared to that of barley. In fact, most part of barley soluble fibre consists in β-glucans and pentosans, and average β-glucans content in this cereal is higher than in wheat and triticale ( FEDNA, 2016 ). Therefore, the higher crude fibre content of barley compared to wheat and triticale (Table 1; FEDNA, 2016 ) may partially explain the lower barley voluntary intake and preference (Table 2) recorded in the present trial, compared to the remaining tested cereals.

Regarding hardness of the cereal kernel, durum wheat has been considered the cereal of highest hardness value, producing coarse and vitreous particles when milling. On the contrary, a soft kernel produces a floury, opaque and fine particle when milling ( Evers & Millar, 2002 ). Wheat hardness depends on the presence of a compound named puroindoline and polar lipids on the starch granule surface ( Pauly et al. , 2013 ). Barley and bread wheat must have an intermediate hardness, while triticale is considered a soft grain with approximately half of the kernel hardness of wheat and barley ( Van Barneveld, 2002 ). According to our results, kernel hardness did not influence the intake by the quails. In fact, the preference test showed that durum wheat was, by far, the most consumed cereal. Although unusual in animal feeding, a study by a Canadian group proved durum wheat to be a good grain to feed broilers, with a high apparent metabolisable energy and a low digesta viscosity ( Silversides, 1999 ). As summarised by Sarica et al. (2009) , wheat has a limited use in commercial quail and poultry diets because of its content of soluble non-starch polysaccharides (predominantly arabinoxylans) in the endosperm cell wall ( Mathlouthi et al. , 2003 ). It is known that the water-soluble arabinoxylans of wheat, and the β-glucans and pentosans of barley binds variable amounts of water and increases the digesta viscosity in the small intestine ( Salobir et al. , 1995 ; FEDNA, 2016 ). The presence of these compounds reduces voluntary intake and nutrient digestion in the foregut by slowing the passage of digested nutrients to the gut wall and exposure time of digesta to digestive enzymes ( Sarica et al. , 2009 ; FEDNA, 2016 ). For this reason, the inclusion of these cereals in poultry feed is limited to 20-30% for wheat, 30-40% for triticale and 25-45% for barley, depending on poultry production type (broilers, layers, breeders; FEDNA, 2016 ).

Despite its anti-nutritional factors content, several authors ( Sethi et al. , 2006 ; Yasar & Gok, 2014 ) observed that feed intake of quails fed diet with partial replacement of corn by wheat was not reduced in comparison to control corn-based diets. This fin­ding explains the relatively good performance of wheat grain when used in quail feed formulation and manufacturing and it agrees with the higher voluntary intake (Table 2) and preference (at least for the durum wheat; Table 2) found in the present research. In fact, wheat is considered a palatable cereal in all species ( FEDNA, 2016 ).

The fact that quails consumed a lower amount of barley and the highest amount of durum wheat compared to other cereals, can be explained because quails, like most poultry species, can match their protein intake closely to their requirements when supplied two feeds to choose between them. This has been observed in growing quails by Canoğullari et al. (2004) when birds received ground wheat and concentrate feed simultaneously as a choice, and by McLeod & Dabutha (1997) in birds allowed to choose between a low-energy, soya-based, high protein mixture and a high-energy, wheat-based, low protein mixture under several ambient temperatures. Toprak & Yilmaz (2012) also reported feed intake reduction in quails fed barley-based deficient diets. In general, cereal consumption decreased as protein intake decreased. In fact, protein content of barley was the lowest of the tested cereals, while the one of durum wheat was the highest (Table 1).

Except for barley, cereal daily intake recorded in the preference trial for each single grain, particularly durum wheat, was near to that recorded for wheat (6.03 g/d) in an experiment carried out by Canoğullari et al. (2004) on growing quails when offered ground wheat and concentrate feed simultaneously as a choice. Moreover, choice feeding did not change feed intake when total daily intake of the four whole grain cereal supplied together recorded in the preference trial (14.3±0.21 g) was compared to daily intake of each single cereal when available alone (12.1 to 15.6 g; Table 2). This result also agrees with findings by Canoğullari et al. (2004) when offering ground wheat and concentrate feed simultaneously to growing quails, compared to supplying only concentrate.

Although adult female Japanese quails are heavier than males ( Vali, 2009 ), in the present study no differences between sexes were found in relation to cereal intake and preference when sex was considered as a factor because bird weight was included as a covariate in the statistical analysis. This fact agrees with previous findings reporting similar feed intake in quails of both sexes ( Vali, 2009 ; Chin et al. , 2013 ). However, females are more sensitive than males to undernutrition ( Chin et al. , 2013 ), something that could explain the interactions between sex and feedstuff factors found in the present research (Table 2), in the sense that: i) balanced feed intake was higher for females than males; ii) males preferred barley in a higher degree compared to females. This could be explained because barley has the lowest nutritive value compared to the other tested cereals, and the cereals have lower (unbalan­ced) nutritive value than the balanced feed.

Positive correlations found, in general terms, between bird live weight and feed consumption in the intake trial (Table 3) agree with previous studies reporting increasing feed intake with increasing body weight ( Marks, 1993 ; Vali, 2009 ), because maintenan­ce needs grow as bird body mass increases ( Marks, 1991 ). The weak (or lack of) correlation between quail live weight and each cereal consumption found in the preference trial (Table 3) might be due to the fact that the four tested cereals were offered simultaneously, thus leading to a greatly different single cereal con­sumptions by each bird. Therefore, birds seemed to be able to adjust and counterbalance their total intake by selecting and combining different amounts of each single cereal, as a way to equilibrate total intake to better fit the maintenance needs ( Pousga et al. , 2005 ). This also explains the strong positive linear correla­tion between quail live weight and the total intake when the four cereals where available simultaneously in the preference trial.

This research may contribute useful information to feed manufacturing and management in production systems in which the use of whole-grain cereals is common. Thus, in rural small-scale poultry production the use of whole grain not only save grinding and mixing cost but it has also demonstrated increased efficiency of diet utilisation ( Pousga et al. , 2005 ). Whole-grain cereals are also given to quails in the finishing phase of the game farming of this species, with the aim of preparing birds for the transition from the balanced feed that receive in the farm to the food that will eat in the wild, where they may also find cereal grains ( Dalmau, 1994 ).

In summary, results from the two trials in the cur­rent study suggest that diet formulation and manufac­turing with whole-grain cereals taking into account preferences of Japanese quails might have nutritional relevance. Japanese quails in these trials chose those cereals that better fit their nutritional requirements. Vo­luntary intake of durum and bread wheat and triticale was higher than that of barley, and intake of each cereal grain was lower compared to the balanced feed. When given as a choice, quails preferred, in decreasing order, durum wheat, triticale, bread wheat, and barley. These differences in voluntary intake and preference among whole-grain cereals, that did not show any differences between sexes, were attributable to differences in their chemical composition, physical texture, and nutritive value.

The authors thank Dr. Ignacio Solís for providing the cereal samples and Dr. José Luis Guzmán for performing the feedstuffs proximate composition ana­lyses. Starch and mineral analyses of feedstuffs were carried out at the Servicio General de Investigación Agraria of the University of Seville (Spain).