Environmental effects of excess P from manure and the soaring price of
   phosphates are major issues in pig production. To optimize P
   utilization, it is crucial to improve our capacity to predict the amount
   of P absorbed, while taking into account the main factors of variation.
   Mathematical modeling can represent the complexity of the processes and
   interactions in determining the digestive utilization of P in growing
   pigs. This paper describes and evaluates a model developed to simulate
   the fate of the dietary forms of P in the digestive tract of growing
   pigs, with particular emphasis on the effect of dietary Ca and exogenous
   phytase on P digestive utilization. The model consists of 3 compartments
   associated with specific anatomical sections: stomach, proximal small
   intestine, and distal small intestine. The main metabolic processes
   occurring in these sections are, respectively, P
   solubilization/insolubilization and phytate P hydrolysis, and P
   absorption and P insolubilization. Model parameters governing these
   flows were derived from in vitro and in vivo literature data. The
   sensitivity analysis revealed that the model was stable within a large
   range of model parameter values (+/- 1.5 SD). The model was able to
   predict the efficacy of Aspergillus niger microbial phytase in
   accordance with literature values, as well as the decreased efficacy of
   plant phytase compared with microbial phytase. The prediction
   capabilities of the model were assessed by comparing actual and
   simulated P and Ca apparent total-tract digestibility (ATTD) based on
   published pig data not used for model development. Prediction of P
   digestibility across 66 experiments and 281 observations was adequate [P
   ATTD observed = 0.24 (SE, 0.943) + 0.98 (SE, 0.0196) x P ATTD predicted;
   R(2), 0.90; disturbance error (ED), 96.5%], whereas prediction of Ca
   digestibility across 47 experiments and 193 observations was less
   accurate (Ca ATTD observed = 11.1 + 0.75 x Ca ATTD predicted; R(2),
   0.78; ED, 20.4%). A lack of agreement between experimental and simulated
   Ca digestibility was found. This model is, therefore, useful in
   evaluating P digestibility for different feedstuffs and feeding
   strategies. It can also be used to provide insight for improving dietary
   P utilization, especially from plant sources, by quantifying the effect
   of the mean sources of variation affecting P utilization.