Deliverable D3.3 A simulation model to predict the post-digestion nutrient use in monogastric animals

Objectives

The objective of the work was to provide the simulation model to predict post-digestive nutrient use in monogastric animals, particularly pigs and poultry. InraPorc was used as the starting point and was extended with an optimization module to calibrate the input parameters according to on-farm data. Different modules have been integrated in the model:

• P-module: simulating the P and Ca metabolism
• FI-module: estimating the feed intake according to environmental conditions (dietary factors, ambient temperature and social environment)
• FA-module: predicting the fatty acid composition of the body and valuable body parts.

The pig model has been adapted to chicken. The nutrient partitioning model for pigs and chicken simulates the post-digestive utilization of energy and amino acids, as well as the Ca and P metabolism and predicts the growth performance and chemical body composition of the individual animals and birds in time.

Rationale

There are number of models predicting energy and protein utilization in pigs and poultry. Despite differences, monogastric species share many similarities in their digestive and metabolic processes. For this reason, there are benefits from developing a common platform that is able to model these processes from a generic perspective, before developing species or system-specific models. A precise and well-defined model for growing animals can be used as a starting point for developing the generic model. For that reason, InraPorc seemed to be sufficient to represent the energy and protein utilization. Considering that the stoichiometry of the underlying metabolic pathways is independent of the species, the common basis and the generic approach seems to be feasible. To be able to initiate the work with InraPorc, the model was reprogrammed in the MATLAB environment that is suitable for extending and improving the model, as well as for further programming both in WP3 and WP4. The results of the simulations of the MATLAB version are identical to that of the InraPorc software. An optimization module on MATLAB was also programmed to be able to calibrate model parameters (animal profile) according to existing databases or farm data. If one has frequent data of feed intake and body weight, an optimization procedure can be run for each individual to identify the phenotypic parameters (PDmean, Precocity, FI50&FI100 and FI1&FI2, in pigs and poultry, respectively) and perform simulation by InraPorc_MATLAB afterwards.

The InraPorc software simulates the utilization of dietary energy and protein in pigs. The amount of ingested feed per day is estimated based on body weight and phenotypic traits. However, it does not take into account that dietary and environmental factors might limit the actual feed intake. InraPorc does not include equations to predict phosphorus (P) metabolism, nor does it predict the fatty acid composition of the body. These issues are included in the model development of the Feed-a-Gene project. Also, the model has been adapted to poultry. For that purpose, new species-specific parameters and equations have been developed.

The feed intake module takes into account the constrains of the amount of feed ingested by the animals. The capacity of gastro-intestinal tract, the environmental temperature, stocking density and the P-supply are considered as the key factors that may limit feed intake. A P-module was introduced to the energy and protein metabolism model predicting the bone and body P retention at different Ca and P supplies. It is a comprehensive description of the underlying mechanism of P utilization in growing and fattening pigs. The input parameters are the nutrient content of the feed, particularly dry matter, dietary Ca and P, as well as Ca and P digestibility. The equations are integrated into InraPorc_MATLAB program that is considered as a starting point for modelling the feed use mechanisms. The P-module presents the distribution of absorbed P and Ca in the body. Absorbed P and Ca are used for maintenance purposes, soft tissue (muscle and backfat), and bone tissue development. Surplus P and Ca are excreted via urine. Retention of P in the body is the sum of P retention in soft and bone tissues. Thus, the model is able to predict P-retention, urinary P excretion and digestible P requirement of swine at different body weights and upon different P supplies. The fatty acid composition is predicted based on ingested dietary fatty acids. The model accounts for the de novo fatty acid synthesis and for the main fatty acid metabolic transactions and simulates the fatty acid composition of the deposited fat tissues in different parts of the pig body.

The pig model represents the nutrient partitioning with a generic approach, thus the core of the model was able to be used for simulating the underlying mechanism of growth in poultry. Modules simulating the energy and amino acid metabolism, as well as the bone mineralization have been modified by using poultry specific parameters taken directly or calculated from the literature. In case of feather production, new equations have been integrated in the model. The poultry model also contains a more detailed feed intake module. However, fatty acid composition of the body is not predicted for poultry.