| Abstrakt: |
High litter moisture content, often referred to as wet litter, is a major problem in poultry production. Wet litter is often related to poor management, diseases, and digestive problems. In this experiment, the objective was to study the relationship between nutrient content and the moisture content of the excreta of broilers. A dataset containing 351 observations was built and contained the nutrient contents data including moisture content of excreta samples collected in 8 different broiler feeding trials. A biological based model approach was used to create a model with 10 and another one with 14 variables that may explain the excreta moisture level response. Subsequently, these models were compared with a statistical model that was built automatically and adjusted only if this improved the biological model. The R(2) of the 10 variable model was 0.54, in which Zn content and the interaction of NDF × K and Ca × P content were negatively associated with excreta moisture. Sodium, P, and Ca content and the interaction between content of NDF × Na were positively associated with excreta moisture. The R(2) of the 14 variable model was 0.58, in which Zn and K content and the interaction of NDF × protein and Ca × P content were negatively associated with excreta moisture, and Na, protein, P, and Ca content and the interactions in contents of NDF × Na, NDF × Zn, and K × Cu were positively associated with excreta moisture content. In conclusion, the models confirmed the effect of Na, protein, P, and Ca on excreta moisture content. Furthermore, hitherto unknown nutrient interactions that contribute to excreta moisture level were identified. As excreta levels of most nutrients can be manipulated by adjusting dietary nutrient levels, dietary formulation can be adjusted with the findings of this analysis to change levels of excreted nutrients and, consequently, also moisture output. |
