The high fibre content of wheat bran limits its use to herbivorous and omnivorous fish. General recommended rates are 2-5 % and wheat bran should preferably be extruded (Hertrampf et al., 2000). Wheat bran has been tested and used in numerous fish species and higher inclusion rates seem possible in certain cases.
Nile tilapia (Oreochromis niloticus)
The nutrient digestibility of wheat bran in Nile tilapia was found to be relatively high for protein (75-84 %) and amino acids (78-87 %) but very low for energy (37-39 %) and generally much lower than nutrient digestibilities of fish meal and other protein sources (Maina et al., 2002; Ribeiro et al., 2011; Sklan et al., 2004). Wheat bran is one of the main ingredients used by tilapia farmers in Subsaharan Africa (El-Sayed, 2013). In Kenya, with Nile tilapia fingerlings fed cereal brans (maize, wheat and rice) at 1.5 % body weight, growth obtained with the wheat bran treatment was intermediate between that allowed by maize bran (highest) and rice bran (lowest). However, wheat bran was more profitable (Liti et al., 2006).
Common carp (Cyprinus carpio L.)
Common carps fed wheat bran at 3 % body weight 3 times daily had better growth, feed conversion, DM in carcass, protein and energy retention than carps fed rice bran (extruded or not) but performance were lower than for carps fed biscuit wastes and wheat middlings (Shalaby et al., 1989). In carps fed a diet containing 48.5 % of test ingredient, protein digestibility was higher for wheat bran (81 %) than for sorghum grain but lower than for rye. Lipid digestibility was the highest for wheat bran (82 %) (Degani, 2006). A carp diet containing 10 % of wheat bran gave better peformance than when 20 % wheat bran was used (Rahman et al., 1989, cited by Hertrampf et al., 2000).
Other carp species
Wheat bran included at 30 % in the diets of fingerlings of 3 Indian carp species (Catla catla, Labeo rohita, Cirrhinus mrigala) was found to be a suitable ingredient, with relatively high digestibilities for protein (93 %) and energy (80 %). Carbohydrate digestibility was lower (68 %) and inferior to that of cooked maize but still higher than that of rice bran and rice polishings (Erfanullah et al., 1998). Labeo rohita carps fed diets containing up to 5 to 15 % wheat bran had a better weight gain during the first 3 weeks at the 15 % rate while weight gains were identical after 3 weeks. Phosphorus retention was decreased in the groups given 10 and 15% bran (Mitra, 1988). In Cirrhinus mrigala fingerlings fed ingredients at 4 % of wet body weight twice a day, wheat bran gave lower body weight and higher feed conversion ratio than maize guten feed and sunflower meal (Shabir et al., 2003).
In South American characids black pacu (tambaqui, gamitana) Colossoma macropomum and red pacu Piaractus brachypomus, cassava root, plantain fruit and peach-palm fruit (Bactris gasipaes) gave better growth performance than wheat bran and wheat middlings in diets containing 30 % of the test ingredient (Lochmann et al., 2009). In Colossoma macropomum, a diet including 23-25 % wheat bran formulated to containing 25 % protein and a DE of 11.3 MJ/kg gave better growth and feed conversion than diets containing 35 % protein, or than diets containing 25 % protein and lower DE (Gutierrez et al., 2009). Dry matter and crude protein digestibilities in red pacu Piaractus brachypomus were found to be lower for wheat bran than for soybean meal, fish and maize grain and the apparent DE of wheat bran was estimated at 7.5 MJ/kg, half of that of maize (14.0 MJ/kg), 2/3 that of soybean meal (10.0 MJ/kg) and much lower than that of fish meal (16.0 MJ/kg) (Fernandes et al., 2004).
In juvenile cichlids Sarotherodon melanotheron, maize bran was more suitable feed than chicken droppings and wheat bran. Survival rate with the by-products was always lower than with commercial feeds (Ouattara et al., 2005).
Mullet (Mugil auratus)
Mullet fry could be reared on a diet containing fish meal and wheat bran at the ration of 3:1 and fed at 10 % of body weight (Enbayah et al., 1987 cited by Hertrampf et al., 1989).