Effect of pellet-size on fish growth, feeding behaviour and natural food web in pond polyculture

In carp-tilapia pond polyculture, carps are stocked once and usually harvested each 12 to 24 months, while, tilapia is stocked and harvested each 4 to 6 months. Therefore, per carp cycle there are 3 to 6 tilapia cycles. With time, carps grow bigger while tilapia reach only 10 to 50% the mass of carps. Pellet size can offer a feeding management tool in the ponds with large carp and small tilapia. This experiment investigated whether pellet-size can steer the growth of large carps (∼ 500 g) and small tilapias (∼30 g) in polyculture ponds. Four pellet-size treatments were applied in this study, each with a different combination of small (2 mm) and large (6 mm) pellets (on weight basis): S100L0 (100% small, 0% large), S50L50 (50% small, 50% large), S25L75 (25% small, 75% large) and S0L100 (0% small, 100% large). Each treatment was applied to six replicate ponds. The experiment lasted for 8 weeks. The treatments did not alter carp production, but tilapia production was highest in S100L0 treatment ponds (P < 0.001). The fish always needed more time to consume large pellets than the small pellets. Small pellets were consumed much quicker when the portion fed was small, indicating strong competition. A similar trend was observed with large pellets but the difference in feed intake time across the treatments was smaller than with small pellets. This is because tilapia ate the majority of both small and large pellets while carps ate less pellets. Tilapia at treatment S0L100 had a lower protein and fat content and a higher ash content than at the other pellet-size treatments. Carp's body composition remain unaffected by pellet-size treatment except the ash content of rohu which was highest at S100L0 and lowest at S0L100. By volume, tilapia had a higher amount of food in the stomach in treatments S100 L0 and S50L50 than in the other treatments. In contrast, more food was found in the gut of rohu in treatments S0L100 and S25L75 than in the other treatments. Pellet-size treatments did not affect water quality and the pond food web (P > 0.05) except phytoplankton diversity which was positively influenced when feeding more large pellets. The plankton population increased with the progression of the experiment, suggesting the food web was not fully exploited. In conclusion, pellet-size proves to be a feeding management tool to steer the production of tilapia as well as of total fish biomass, but not of carp's biomass in carp-tilapia pond polyculture.