An integrated in vitro gastrointestinal digestion and fermentation model as a complementary tool for evaluating feed ingredients in young broilers: a case study with peas

The search for sustainable protein sources in broiler nutrition has increased interest in locally produced ingredients such as peas. However, nutritional evaluation of alternative feed ingredients relies on costly and ethically constrained animal trials, while standardized in vitro models representing the complete gastrointestinal tract of young broilers are lacking. These models should reflect their lower digestive capacity and less complex caecal microbiota. This study developed an innovative in vitro model simulating the complete gastrointestinal tract of young broilers and used peas as a case study to evaluate partial replacement of soybean meal and compare in vitro and in vivo responses. Isoenergetic and isonitrogenous feeds were formulated with 0, 10, 20, or 30% pea inclusion. An in vivo trial with 10-day-old broilers assessed growth performance, protein digestibility, intestinal morphology, caecal fermentation metabolites, and microbiota. In parallel, the same feeds were subjected to the in vitro model, including gastric and small-intestinal digestion and fermentation using a broiler caecal inoculum. Increasing pea inclusion reduced growth performance and protein digestibility and was associated with lower ileal villus height, indicating reduced nutrient utilization at higher raw pea inclusion levels. In vivo caecal short-chain fatty acid and ammonia concentrations were not markedly affected, although ammonia showed a numerical decrease. In vitro, feeds containing 20 or 30% peas showed significantly different digestion kinetics. During in vitro fermentation, pea inclusion increased gas production and acetate and butyrate formation, while reducing ammonia production, suggesting more saccharolytic fermentation. While in vitro results agreed with in vivo protein digestion, caecal fermentation responses were more pronounced in vitro. These findings indicate that raw peas can partially replace soybean meal in young broiler diets and that complementary in vitro and in vivo approaches integrate controlled assessment with physiological confirmation in animals. The model offers controlled assessment before targeted in vivo evaluation, although further research should improve its predictive capacity for caecal fermentation by assessing inoculum representativeness and the selection of soluble and fine-particle digesta fractions for fermentation.