Despite the effort that has been made in the development of formulated starter feeds for larval fish, live food still remains a better option in terms of growth and survival compared to formulated diets (Verreth et al., 1987; Kolkovski et al., 1995; Garcia-Ortega et al., 1998; Mitra et al., 2007).

Mehmood et al. (1998) reported that the provision of live food to first feeding rohu larvae enhances growth, health and survival and hence significantly reduces rearing costs. In rohu larviculture, natural live food is provided through pond fertilization schedules. Jana and Chakrabarti (1990) and Chakrabarti and Jana (1998) suggested that exogenous introduction of live zooplankton into larval rearing ponds may be a better approach to rohu and mrigal fry rearing than indirect stimulation of pond plankton production through manuring.

Large commercial scale carp culture has been practiced in India for the past 2-3 decades, particularly in Andhra Pradesh and Punjab. More than 90 percent of farmers use farm-made feeds consisting of an oilseed cake and rice/wheat bran mixture (Mohanty, 2006). The most commonly used farm-made supplemental feed is mixture of rice bran and groundnut oil cake (50:50 ratio) with the average proximate composition of crude protein 24 percent, crude lipid 9 percent, ash 9-10 and digestible carbohydrate 45-48 percent. Supplementary feeds for Indian major carps are manufactured by the cattle and poultry feed industry. In general the feeds have very poor water stability.  The proximate composition of factory made feeds is reported to be 20-30 percent protein, 2-4 percent lipid, 10-15 percent fibre, 30-40 percent carbohydrate and 8-10 percent ash and often are claimed to have been enriched with lysine, methionine, vitamins and minerals (Nandeeshsa, 1993) but because of price commercial feeds are not commonly used for carp culture in India. However, in the state of Punjab as well in other parts of the country, commercially manufactured pellets are now available at a competitive price (Figure 7). During the 1990s less than 8 percent of farmers in Andhra Pradesh used factory made feeds (Verrina et al., 1993) and this situation remains the same today. The major problems with the use of commercial feeds relate to pellet stability and proper assessments of the impact of low cost feeds on growth and economics of production. The advantages and disadvantages of different feeds are presented in Table 8. Since feed contributes about 50-60 percent of total production cost, several non-conventional ingredients have been tried in recent years. Recommended inclusion levels of alternative protein sources are presented in Table 10. Several feed formulations have been developed for different life history stages of Indian major carps, the composition of which are listed in Table 11.

De Silva (1993) questioned the appropriateness of Indian major carp feed formulations that contain a significant proportion of fishmeal and those that have been supplemented with micronutrients. According to De Silva and Gunasekera (1991) and De Silva (1993), in semi-intensive aquaculture where external feed input is expected to supplement natural food production, the use of nutritionally balanced feeds constitute an unwise use of resources and economically unsound practices. Their argument is based principally on the significant differences between the dietary protein requirements for optimal growth and the economically optimum protein level.

During the 15 day fry rearing period supplementary feed is provided at a rate of 400 percent of the initial body biomass for the first week and during the second week feed is provided at 800 percent of fish biomass. The daily ration is either supplied once in the morning or divided into two equal portions and fed to the fish in the morning and in the evening. Biswas et al. (2006a) suggested that feeding the fish once a day is adequate during the nursery period when natural food (plankton) is available in adequate quantities.

During the fingerling rearing period (about 90 days), supplementary feed is normally provided at a ration of 8-10 percent of the initial biomass of fry per day during the first month, and at 6-8 percent of biomass during the subsequent two months. However, Ahmed (2007) suggested that feeding late fry at a rate of 6.5–7.0 percent body weight per day, with a 26-28 percent protein diet is optimum for growth and efficient feed utilization. Under polyculture conditions, Biswas et al. (2006b) found that there is no benefit in feeding late fry more than once a day if there is an adequate abundance of natural live food.

During grow-out, supplementary feed is provided once (or twice) a day at a daily ration of  3-5 percent of the biomass in the first month and then reduced to 1-3 percent of fish biomass per day (Jena and Das, 2006).  Further detail of feeding schedules is provided in Table 12.

Mixed feeding schedule, using a low protein diet (rice bran) alternating with the  high  protein diet (rice bran and ground nut oil cake) in daily feeding was found to have positive effect on growth besides significant savings on feed cost as well as protein. By using different feeding schedules, it was possible to save 15-31 percent protein as well as 10-20 percent feed cost (Nandeesha et al., 1994). Mohanty and Samal (1994) have also observed the benefits of using mixed feeding schedule in rohu. Saha and Ray (1998) also reported that feeding the fish with leaf meal diets on one day, followed by three days of an animal protein based diet resulted in better performance. Studies clearly have demonstrated that feeding fish with diets containing different levels of protein give better results than feeding fish with high protein diet on a regular basis. This strategy is useful not only to save feed cost, but also to reduce environment pollution through discharge of reduced level nutrients.