|dc.description.abstract||A meta-analysis was performed using five random effects models to compare the growth performance of Nile tilapia (Oreochromis niloticus) fed different carbohydrate sources (raw starch versus glucose; raw starch versus dextrin; hydrothermally processed starch versus glucose; hydrothermally processed starch versus dextrin; and starch sources with different amylose content). The standardized mean difference (SMD) of growth indicators was calculated using Hedges’ g (HG). The growth performance of fish fed raw starch in the diet was significantly higher than glucose (SMD = -0.64; P < 0.05), but significantly lower than dextrin (SMD = 0.61; P < 0.05). The starch source with amylose content under 10% was considered as the control group to interpret the effects of different amylose and amylopectin content to growth. There was a significant quadratic relationship between effect size and amylose content in starch source (P < 0.01, r2 = 0.545).
Following the meta-analysis, three experiments were performed to determine the 1) digestibility, 2) growth performance and 3) glycemic indices (GI) in Nile tilapia fed different pulse starches. In Experiment 1, 405 mixed sex Nile tilapia (mean weight 336 g) were fed one of five diets containing 700 g kg-1 of the reference diet and 300 g kg-1 of the experimental ingredient (modified cornstarch, pea, lentil or faba bean starch). The apparent digestibility coefficients (ADCs) for dry matter, crude protein, gross energy and starch varied between the four starch ingredients. The ADC of starch in the four ingredients ranged from 50.5 to 60.9. Lentil starch was significantly higher for all digestibility parameters compared to the other starch sources.
An 8-week growth trial was then conducted where Nile tilapia (mean weight 531 g) were fed diets containing either 0, 150 or 300 g kg-1 modified corn, pea, lentil or faba bean starch (306 g kg-1 digestible crude protein and 3600 kcal kg-1 digestible energy based on the digestibility results from Experiment 1. There was a significant quadratic relationship found between starch inclusion levels and average daily gain (ADG), specific growth rate (SGR), or feed conversion ratio (FCR) in fish fed diets formulated with faba bean starch. The highest growth performance was achieved when faba bean starch inclusion was approximately 200 g kg-1 (P < 0.05), with lower growth performance at inclusion rates both above and below this level. In contrast, no significant relationship was found between faba bean starch inclusion and average daily feed intake (ADFI). A significantly positive linear relationship was seen between the remaining starch inclusion levels and ADG or SGR (P < 0.05). Lastly, a negative linear relationship was found between pea starch or modified cornstarch inclusion and FCR, but no significant difference was found between lentil starch inclusion and FCR.
In Experiment 3, fasted Nile tilapia (6 per ingredient) were anesthetized and force fed 0.5 g available carbohydrate per kg body weight of glucose, pure corn, pea, lentil or faba bean starch. Blood samples were taken from the caudal vein at 0, 3, 6, 12, 24, and 48 h after force-feeding. All starch sources produced a glycemic response that peaked around 12 hours after feeding. Using the glycemic response to pure glucose as a control index at 100, the GI of corn, pea, lentil and faba bean starches were 61, 54, 69, and 76, respectively, with no significant differences found between them.
In conclusion, Nile tilapia was able to digest pulses starch and was able to absorb and effectively handle glucose. These results suggest that pea and lentil starch can be included in Nile tilapia diets up to 300 g kg-1 without negatively affecting fish growth performance but faba bean starch should not be fed above 200 g kg-1. Therefore, pulse starches were recommended in tilapia diets to reduce the use of ecologically and economically costly fishmeal, which contribute to the sustainable development of aquaculture.||