Mathematical modeling of digestion and nutrient absorption in pigs

D. Bastianelli, D. Sauvant and A. Rerat

J ANIM SCI 1996, 74:1873-1887.

The online version of this article, along with updated information and services, is located on
the World Wide Web at:
http://jas.fass.org/content/74/8/1873

www.asas.org

Downloaded from jas.fass.org by guest on July 14, 2011

 Mathematical Modeling of Digestion and Nutrient Absorption in Pigs

Denis Bastianelli*, Daniel Sauvant*,1 and Alain Rérat†

*Laboratoire de Nutrition et Alimentation, Institut National de la Recherche Agronomique, INAPG,

F-75231 Paris cedex 05, France and †Département de Nutrition, Alimentation, Sécurité Alimentaire,
Institut National de la Recherche Agronomique, F-78350 Jouy en Josas, France

ABSTRACT: A simple simulation model of diges- cal integration with a time step of 1 min allows
tion and absorption in pigs was developed. The prediction of kinetic features of digestion phenomena
structure of the model is a set of four anatomical such as absorption patterns and transit flows. First
compartments for DM: stomach, two portions of small validation of the model shows that the global dynamic
intestine, and the large intestine. In each of these behavior of the model is realistic and promising.
anatomical compartments, subcompartments cor- However, some additional factors must be considered
respond to the major biochemical components of feed for improved accuracy, in particular the susceptibility
and their products of degradation. The major degrada- of the feed components to enzymatic degradation. The
tion and absorption events are considered, as well as outputs of such a model could be used as inputs for
the effect of microbial activity in the large intestine. metabolic or growth models running with time steps
The total number of compartments is 44. The numeri- smaller than the 24-h basis often used in nutrition.

Key Words: Dynamic Models, Digestion, Absorption, Pigs

J. Anim. Sci. 1996. 74:1873–1887

Introduction 1991a,b; Pettigrew et al., 1992a). But surprisingly,

the functions of the digestive tract were always
Published experimental data on digestive physiol- aggregated through one coefficient of digestibility for
ogy in pigs increased during the last three decades as the whole ration (Whittemore and Fawcett, 1974) or
a consequence of research efforts to improve the at best for the main constituents (Black et al., 1986,
efficiency of pig production and to find an animal Pettigrew et al., 1992a). As underlined by Gill et al.
model for humans. However, a large part of the (1989), the lack of a model describing digestion and
experimental results cannot be fully appreciated due absorption is the limiting factor in the attempt to
to an obvious lack of integration. Moreover, there model nonruminant metabolism.
seems to be an increasing gap between the knowledge The present work proposes and discusses a basic
acquired through recent research and the classical structure for a mechanistic model of pig digestion. The
concepts used in animal nutrition (digestibility coeffi- idea of using a compartmental model to describe
cients and other aggregated criteria). digestion and transit has been proposed by Tomassone
Mathematical modeling is a useful tool to integrate and Laplace (1973), but the only results published so
dispersed data and concepts and to achieve a compre- far concern the stomach (e.g., Bernier et al., 1988)
hensive view of complex biological systems (Sauvant, and some aspects of protein digestion in the small
1992; Dijkstra and France, 1995). Several models to intestine (Rivest et al., 1994). A small number of the
predict swine growth have been proposed, and these principles of the current approach have been given in
integrate an ever increasing number of factors Bastianelli et al. (1994).
(Whittemore and Fawcett, 1976; Moughan and Smith,
1984; Black et al., 1986). Some models have already
achieved a relatively high degree of mechanistic General Description of the Model
integration of the metabolic processes from absorbed
nutrients to final use (Schulz, 1978; Pomar et al., The diagram of the model is in Figure 1. The
digestive tract is described as a set of four successive
anatomical compartments ( AC) . The feed enters the
1To whom correspondence should be addressed. first compartment, the stomach ( STO) , and outflows
Received August 4, 1995. to the second. The small intestine is divided into two
Accepted March 5, 1996. parts of unequal size: SI1 represents the first portion,

1873

Downloaded from jas.fass.org by guest on July 14, 2011

1874 BASTIANELLI ET AL.

Figure 1. Simple diagram of the model, showing the four anatomical compartments (AC): stomach (STO), two
parts of small intestine (SI1 and SI2), and large intestine (LIC). Biochemical subcompartments (BSC) are nonprotein
nitrogen (NN), protein (PR), pool of amino acids (AA), starch (ST), sugars (SU), digestible cell walls (CW), lipids (CF),
volatile fatty acids (VFA), fatty acids (FA), undigestible cell walls (UF), and minerals (AS). In addition, there is a
microbial subcompartment in LIC (MI). Flows between compartments are represented (solid lines). Other flows are
endogenous secretions (endo) and absorption (abs), represented by broken lines.

the duodenum and the proximal part of jejunum and [SU], digestible [CW] and undigestible [UF] cell
SI2 represents the medial and distal parts of the walls), nitrogenous compounds (proteins [PR], pool of
jejunum and the ileum. The last AC represents the amino acids [AA], nonprotein nitrogen [NN]), fats
sum of the cecum and the large intestine ( LIC) from (lipids [CF], fatty acids [FA], volatile fatty acids
which the feces are excreted. [VFA]), and minerals ( AS) . Lastly, there is a
Within each AC, biochemical subcompartments microbial subcompartment in the large intestine
( BSC) of assumed homogenous digestive behavior are ( MI) .
distinguished on the basis of the most widely used The contents of all compartments and flows are
analytical procedures. The BSC considered in the expressed in grams of DM. The total number of
model are carbohydrates (starch [ST], soluble sugars subcompartments is 44. The change in mass in each

Downloaded from jas.fass.org by guest on July 14, 2011

 MODEL OF DIGESTION IN PIGS 1875
compartment is described with a deterministic dy- size (grams of DM) of the meals. The duration of the
namic differential equation. The structure of these meals, set by default to 15 min, can be modified. The
equations is given in Tables 1 and 2. The basic vector of the qualitative description of the feed
principle included in the mathematical relationships corresponds to the proportion (percentage of diet DM)
describing the flows is the linear mass-action law, of the meal in each of the biochemical subcompart-
except when a documented reason suggests another ments of the stomach. For simplification, the potential
type of relationship. The time step used for numerical degradability of the fiber fraction (NDF content) is
integration (Euler’s method) of these differential set to a constant value (70%) in this version of the
equations is the minute. For the simulations program. The moisture content of the feed and the
presented here, Professional Dynamo Plus (Pugh- water consumption are not taken into account because
Roberts, 1986) software was used, but the model could all the calculations are performed on a DM basis.
be developed with any other simulation language as
well. Transit
The voluntary intake rate of pigs is generally very
Parameterization high, and the gastric reservoir has therefore a major
smoothing role in transit variations by stocking
The parameterization used in this paper is given in matter and releasing it slowly. This latter aspect is
Table 3. It refers to a pig of 40 kg live weight, which is important because degradation and absorption are
in the range of most of the published data. Moreover, known to be rapid in the proximal parts of the small
at this weight, the small intestine of pigs is supposed intestine, and thus the flow of digesta from the
to have almost reached its final length and capacity stomach can affect nutrient absorption. The physiolog-
(Laplace, 1970; Low and Zebrowska, 1989; Low, ical regulations of gastric emptying are complex and
1990). are related to particle size, viscosity, osmolarity,
lipids, sugars, fiber, and protein content (Darcy, 1984;
Input Parameters Low, 1990). In the study by Rainbird (1986), it
appears that liquid meals are emptied more rapidly
The input to the model is the quantitative and and that factors increasing viscosity (e.g., guar gum)
qualitative description of the feed. The feeding pattern delay the emptying of such meals. However, when
is controlled by the frequency (number per day) and considering mixed feeds with no extreme value for the

Table 1. Notation used in the equations

Ai,c absorption flow of constituent i in AC c

AL*,STO alimentary flow of DM in STO
ALi,STO alimentary flow of constituent i in STO
Dji,c degradation flow of a constituent j in constituent i in AC c
Du duration of a meal
Ei,c endogenous flow of constituent i in AC c
F*,bc flow of DM from AC b to AC c
F*,FEC fecal flow of DM
Fi,bc flow of constituent i from AC b to AC c
Fi,FEC fecal flow of constituent i
GMIC growth of microbial DM
KAi,c Michaelis parameter for absorption flow of constituent i in AC c
KLIC steepness parameter for LIC emptying
Nm number of meals
Pi,AL proportion of constituent i in the feed
Pi,c proportion of constituent i in AC c
Q*,c quantity of DM in AC c
Qal total daily quantity of feed
Qi,c quantity of constituent i in AC c
QNLIC neutral quantity for LIC
RDji,c fractional rate of degradation flow of a constituent i in constituent j in AC c
REi,c fractional rate of endogenous flow of constituent i in AC c
RF*,bc fractional rate of flow of DM from AC b to AC c
RF*,FEC fractional rate of fecal flow of DM
RUi,LIC fractional rate of uptake of constituent i for microbial growth in LIC
Ui,LIC uptake of constituent i for microbial growth in LIC
VAi,c maximum velocity for absorption flow of constituent i in AC c
YG efficiency of microbial growth with respect to glucose

Downloaded from jas.fass.org by guest on July 14, 2011

1876 BASTIANELLI ET AL.

Table 2. General form of the equations

Stomach (STO)
Differential equation:
dQi,STO/dt = ALi,STO + Ei,STO − Fi,STO.SI1
Calculation of flows:
ALi,STO = AL*,STO Pi,AL
AL*,STO = Qal/(Nm*Du) if t e[k*1440/Nm, k*1440/Nm+Du]; 0 at other times
Ei,STO = REi,STO Q*,STO
Fi,STO.SI1 = F*,STO.SI1 Pi,STO
Pi,STO = Qi,STO/Q*,STO
F*,STO.SI1 = RF*,STO.SI1 Q*,STO
Small intestine (SI1 and SI2)
( c is the AC considered, b the previous one and d the next one)
Differential equation:
dQi,c/dt = Fi,bc + Dji,c + Ei,c − ( F i,cd + Dij,c + Ai,c)
Calculation of flows:
Fi,bc = F*,bc Pi,b
Pi,b = Qi,b/Q*,b
F*,bc = RF*,bc Q*,b
Dji,c = RDji,c Qi,c
Ei,c = REji,c F*,bc
Ai,c = VAi,c Qi,c/(Q i,c + KAi,c)
Large intestine (LIC)
Differential equation:
dQi,LIC/dt = Fi,SI2.LIC + Dji,LIC + Ei,LIC − ( F i,.FEC + Dij,LIC + Ai,LIC + Ui,LIC)
dQMIC/dt = GMIC − FMIC,.FEC
Calculation of flows:
Fi,SI2.LIC = F*,SI2.LIC Pi,SI2
Pi,SI2 = Qi,SI2/Q*,SI2
F*,SI2.LIC = RF*,SI2.LIC Q*,SI2
Dji,LIC = RDji,LIC Qi,LIC
Ei,LIC = REji,LIC F*,SI2.LIC
Fj,FEC = F*,FEC Pi,LIC
F*,FEC = RF*,FEC exp((Q *,LIC − QNLIC)/K LIC) Q*,LIC
Ai,LIC = VAi,LIC Qi,LIC/(Q i,LIC + KAi,LIC)
Uj,LIC = GMIC RUi,LIC
GMIC = YG QSU,LIC

previous characteristics, a large part of these factors the patterns of DM and N emptying are similar to
can be neglected in a first approximation. To assess those in Low et al. (1985). This has also been
the validity of this hypothesis, data from publications reported in humans for the emptying of DM, protein,
giving the kinetics of gastric emptying of DM, and lipid, and carbohydrate of a homogeneous meal
sometimes nitrogen, starch, or glucose, have been (Miller et al., 1978). Considering all the above
collected. This database covers regimens from a mentioned aspects, we decided to adopt for the model
protein-free synthetic diet (Rérat and Lougnon, 1963) the value of the slope of Figure 2 as the fractional rate
to an all-cereal diet (Laplace et al., 1985) and feed of flow from the STO.
intakes ranging from 500 g (Rainbird, 1986) to 1,440 The size of SI1 (duodenum + proximal jejunum) is
g DM (Zebrowska and Horszczaruk, 1975). The data set to 15% of the mean residence time of the small
of kinetics of DM emptying are shown in Figure 2: the intestine. It was separated from the rest of the small
intestine because of its particular importance in
natural logarithm of the percentage of initial DM
hydrolysis and absorption (Miller et al., 1978; Bernier
remaining in the stomach is plotted against time after
et al., 1988). Data are scarce on the dynamic aspects
meal. This figure shows a fairly satisfactory agree-
of the digesta flow through the duodenum. Digesta are
ment with the hypothesis of a first-order process. propulsed by peristaltic movements, and SI1 is a
However, data points in the early stage are below the priori not comparable to a simple mixing and delaying
regression line, suggesting that the first phase of reservoir as the stomach. But in a first approximation
gastric emptying is slightly more rapid, as reported by we decided to treat it as a compartment, as did
Auffray et al. (1967). Figure 3 compares the slopes of Bernier et al. (1988). The DM is therefore assumed to
gastric DM, nitrogen, starch, and glucose emptying. leave SI1 following a mass action law. As in other
The nitrogen and starch parallel the DM kinetics, anatomical compartments, each biochemical subcom-
even if glucose emptying seems to be slower, probably partment in the duodenum is assumed to follow the
because of a limited breakdown of starch. In addition, same kinetics of transit as the whole DM.

Downloaded from jas.fass.org by guest on July 14, 2011

 MODEL OF DIGESTION IN PIGS 1877

Figure 2. Gastric emptying of dry matter: natural Figure 3. Gastric emptying of various nutrients:
logarithm of the percentage of DM remaining in the natural logarithm of the percentage of each nutrient
stomach. Data from Cuber and Laplace (1979), Cuber et remaining in the stomach. Data from Cuber et al. (1981,
al. (1980, 1981), Laplace and Cuber (1984), Low et al. 1981), Laplace and Cuber (1984), and Rérat and Lougnon
(1985), Rainbird and Low (1986), Rérat and Lougnon (1963) (o, — −, glucose; ÿ, ——, nitrogen; ∫, − −, starch).
(1963), and Zebrowska and Horszczaruk (1975).

time to pass through SI2. In contrast, no delay was

The second part of the small intestine, jejunum + considered for the digestive events (degradation and
ileum (SI2), is a long tube of approximately 15 m absorption) occurring in SI2, which start as soon as
(Vodovar et al., 1964), and the approximation of a the digesta enters the compartment.
simple mixing compartment is hardly sustainable. In pigs, the cecum is fairly small compared with the
However, we decided to represent it as a single colon. Moreover, even if the characteristics of the
compartment, but with introduction of a delay of order fermentation are not strictly the same as in the colon
3 and mean duration of 60 min for the outflow of DM. (Argenzio, 1982), they are comparable in a first
In the programming language used (Professional approximation (Bach Knudsen et al., 1991). Conse-
Dynamo Plus; Pugh-Roberts, 1986), a delay of order n quently, many studies consider them together when
and duration DUR represents the passage through n analyzing the fermentation processes in pigs. For
virtual compartments (and therefore a smoothing of these reasons, they were pooled into a single AC
the passing flow), with a mean time of DUR between (LIC) in the model. The DM in this compartment is
the original variable and the delayed variable). This the sum of the digesta’s DM and the bacterial DM.
takes into account the fact that the digesta takes some The emptying of the large intestine is considered as a

Table 3. Values of parameters in the simulations presented in this article

ST SU CW PR AA NN CF FA VFA AS
Fractional degradation rates, %/mn
SI1 8 — — 3 — — 2 — — —
SI2 4 — — 2 — — 1 — — —
LIC 2 — .08 .1 — — 0 — — —
Absorption parameters: VA, g/mn; KA, g
SI1 VA — 1 — — .1 0 — .2 — —
KA — 2 — — 2 0 — 1 — —
SI2 VA — 1 — — .2 .2 — .5 — .2
KA — 10 — — 5 5 — 4 — 10
LIC VA — 0 — — — .04 — .05 .2 .03
KA — 0 — — — 10 — 20 10 50
Endogenous secretions, g/kg DM flow
STO — — — 3.9 — 1.3 — — — 6.25
SI1 — — — 12 2.4 5 20 — — 30
SI2 — — — 22 — 10.7 — — — 20
LIC — — — 15 — 4 — — — —
Microbial uptake, g/g microbial DM growth
LIC — 3.33 — — .11 .124 — .063 — 0.13

Downloaded from jas.fass.org by guest on July 14, 2011

1878 BASTIANELLI ET AL.

function of total DM in LIC. But because it is well lipids can be degraded into AA, SU, and FA, respec-
documented that a higher ileal DM flow decreases the tively.
residence time in the large intestine (review of The AA, SU, and FA can be absorbed in small
Warner, 1981), the fractional rate of LIC emptying is intestine as well as AS and NN. In the large intestine,
an exponential rather than a linear function of the the microbial activity becomes the major digestion
large intestinal DM. However, the mathematical form phenomenon, whereas in the previous AC, the quan-
chosen for the function is conceptual rather than titative importance of fermentation was assumed to be
directly derived from a specific experiment: negligible, keeping in mind that the model is not
supposed to apply to very young animals.
dQLI/dt = FLIE × exp((QLI − QN)/Kn) × QLI It is considered that gastric hydrolysis does not
provide significant amounts of end products. This is
with FLIE being the basic fractional rate (constant), generally admitted in the literature (review of Rérat,
QN a “neutral” quantity of DM for which the 1981a), except in the experiments of Keys and de
fractional rate of emptying equals FLIE, and Kn a Barthe (1974b). There is some fermentation in the
constant determining the steepness of the exponential stomach, providing lactic acid and VFA, as shown for
function. Moran (1982) stated that the rectum example by Clemens et al. (1975). However, this
accrues digesta from the LIC to a critical mass and phenomenon was not considered in this model, be-
then voids it. The addition of a rectal AC in the model cause quantities of VFA in the stomach are considera-
would permit simulation of this, but it has no bly less than those in the LIC (Imoto and Namioka,
particular interest apart from giving a more biological 1978a).
appearance to the outflow kinetics. So the outflow of It is considered that the hydrolysis of the feed in the
LIC is continuous in the model, which also facilitates small intestine is not limited by enzyme availability.
some calculations (digestibility coefficients, transit This is suggested by much research for protein
time). degradation (Rérat, 1981a) and for carbohydrate
degradation. Rérat (1981a) mentions that a 90%
Endogenous Secretions pancreatectomy in humans leaves enough enzymatic
activity to perform a normal breakdown.
The input of endogenous secretions in the digestive
Degradation and Absorption of Nitrogenous Com-
tract was considered. The values adopted in the model
pounds in the Small Intestine. The parameters of the
are mainly those of the review of Juste (1982). As
equations describing the degradation of PR and the
suggested by much research (e.g., Corring et al., 1972;
absorption of AA are shown in Table 3. They are not
Gaudichon, 1994), the level of the endogenous input
the same in SI1 and SI2 even if the equations have the
in a compartment is a function of the passage of DM
same form. We presumed that the degradation flows
and is calculated by multiplying the flow of DM
are proportional to the level of the substrate. Only
entering the compartment by a constant. The values of
individual amino acids are considered in the AA
the constants are calculated in such a way that the
compartment, which is a simplification because it is
endogenous input equals the mean of the literature
known that the degradation of proteins produces
data when the segmental digestibility is the one
polypeptides first, then oligopeptides, and then di- or
observed for a standard diet. These “neutral” digesti-
tripeptides that can be absorbed. This last phenome-
bility values are 0% at the entry of duodenum, 20% at
the entry of jejunum, and 70% at the ileum (mean of non is not ignored, but pooled with the absorption of
the 41 values, Noblet et al., 1989). In this way, any individual amino acids. Besides, the amino acids
increase in DM transiting the digestive tract, either absorbed as di- or tripeptides seem to be mainly
because of an increased feed intake or due to a released in the blood as free amino acids.
decreased digestibility (e.g., higher fiber content), The absorption of AA is described as a saturable
increases the endogenous input. function of the quantity of free AA in the lumen, as
suggested by the results of Adibi (1969) and the
Degradation and Absorption Processes review of Matthews (1991). It is based on the
quantity of AA available for absorption, according to
General Hypotheses. The description of the biochem- Bernier et al. (1988), and to the constant absorption
ical subcompartments is achieved at a fairly ag- of AA (as well as of lipids and carbohydrates) from
gregated level: only the polymers and the individual the duodenum of humans, as long as the quantity of
absorbable end products are considered. It is assumed nutrients remains high (Miller et al., 1978). However,
that the degradations that occur in the mouth and in the total quantity of AA absorbed is the consequence
the stomach are partial and provide only negligible of the absorption of all individual AA and di- and
quantities of absorbable end products. Therefore, no tripeptides. Thus, it is not possible to extrapolate the
degradation is taken into account before the small results on individual amino acids, such as those of
intestine. Moreover, the stomach absorption is Adibi (1969). For this reason, the concept of the
neglected. In the small intestine, protein, starch, and Michaelian process is used to provide a saturable

Downloaded from jas.fass.org by guest on July 14, 2011

 MODEL OF DIGESTION IN PIGS 1879
function, and the parameters are adjusted to data on The structure is simplified to be consistent with the
global absorption rates of amino nitrogen in the level of aggregation of the other parts of the model.
duodenum or jejunum (Nixon and Mawer, 1970; The bacteria are a subcompartment of LIC with a
Miller et al., 1978). As some NN enters the stomach fixed chemical composition as in most of the ruminal
and small intestine via endogenous secretions, there is models published. Because of the lack of information
a NN subcompartment in these AC and hence a on the composition of pig colonic microflora and
potential NN absorption flow. This last function is because of the degree of similarity between ruminal
difficult to quantify, because it seems to be related to and colonic bacteria, the composition is assumed to be
uremia and consequently to the nitrogen status of the that of ruminal bacteria. According to Merry and
animal (Rérat, 1981b), which is not considered in the McAllan (1983), Storm and Orskov (1983), and
model. The NN absorption is considered as a simple Czerkawski (1976), mean values are (percentage of
Michaelis function of the quantity present in each AC, DM) protein 55%, fat 12.5%, carbohydrates 20%, and
and the parameters are adjusted empirically. minerals 12.5%. As it seems likely that no significant
Degradation and Absorption of Carbohydrates in the degradation due to animal enzymes occur in the LIC,
Small Intestine. The general principles applied for the all degradation occurring in this compartment is
degradation and absorption of carbohydrates are assumed to be due to the microbes. Proteins reaching
similar to those applied for nitrogenous compounds. the LIC can be degraded into AA. Starch and
The degradation of CW is neglected (e.g., Keys and de digestible fibers are degraded into SU. Amino acids
Barthe, 1974a; Vervaeke et al., 1989). The degrada- can be deaminated into SU + NN (Mason, 1984) in
tion of starch is assumed to be proportional to the stoichiometric proportions (.84 and .16, respectively).
quantity in each BSC, the fractional degradation rate Fat can be hydrolyzed into FA. All the degradation
being a constant. The reaction considered is the flows are proportional to the substrate in LIC. The
degradation of ST to SU; the intermediary polyosides values of the degradation rates are shown in Table 3.
of this degradation are included in the ST BSC as Carbon flow from SU is used in a fixed proportion for
nonabsorbable products. the bacterial growth and for the production of VFA.
The absorption of SU is supposed to be saturable The value of this proportion used for bacterial growth
and therefore is described as a Michaelis function. The
is 30%, as suggested by the mean efficiency of digested
values of the parameters are taken from Bernier et al.
OM for protein growth found in the literature (e.g.,
(1988) and from the calculation of the global disap-
Dierick et al., 1990). Bacterial growth consumes other
pearance of glucose from the duodenum or the
nutrients according to their proportion in the bacteria
jejunum (Miller et al., 1978).
and to the yield of biochemical pathways. According to
Degradation and Absorption of Fat in the Small
the results of Giusi (1986), it is assumed that the
Intestine. Fats are assumed to be mainly in the
molar proportions of VFA produced are 70, 20, and 10
triglyceride form and to be hydrolyzed into fatty acids
for acetate, propionate, and butyrate, respectively.
and monoglycerides ( F A ) before being absorbed
Under this hypothesis, the stoichiometry of VFA
(Clément, 1980; Friedman and Nylund, 1980). The
production (Hungate, 1966) predicts the quantity of
rate of lipid absorption is high (Sambrook, 1979a) but
gas (CH 4 and CO2) produced. Soluble sugars are
less than for carbohydrates (Miller et al., 1978).
assumed to have a negligible and constant concentra-
Absorption of Minerals in the Small Intestine. The
tion (Sambrook, 1979b). There is no absorption of SU
aim of the present model is more related to the energy-
in the LIC. Therefore, the prediction of SU released by
yielding and nitrogenous nutrients. Absorption of
substrate degradation and SU used for microbial
minerals is a complex matter, with much reference to
growth allows the calculation of VFA produced by
the hydromineral balance of the intestines and in the
whole animal. It seems therefore impossible to build a difference. The basic hypothesis is that energy is
simple model, particularly with our choice of consider- always the limiting factor in the LIC microbial
ing only DM. Therefore, the model remains very metabolism. The quantity of nitrogen is supposed to be
aggregative at this level and does not reflect the sufficient, because the diffusion of urea of endogenous
mechanisms involved. The first-order processes driv- origin through the LIC wall permits this situation.
ing the absorption of minerals are empirically ad- The presence of NN in sufficient amounts for bacterial
justed on ileal digestibility values of the literature. growth is ensured in the model by a fairly constant
secretion (indexed on the DM quantity like other
Fermentation and Absorption endogenous secretions) combined with a Michaelis
in the Large Intestine absorption. At the LIC level, there is no absorption of
AA (Darragh et al., 1994), but NN, VFA, and AS can
The LIC compartment is modeled as a fermentor, be absorbed. As in the small intestine, AS absorption
on the example of ruminal or RUSITEC models is adjusted empirically on the differences between ileal
already published or developed in our laboratory and fecal digestibility values, and the other nutrients
(Sauvant and Giger-Reverdin, unpublished results). are absorbed according to Michaelis kinetics.

Downloaded from jas.fass.org by guest on July 14, 2011

1880 BASTIANELLI ET AL.

Table 4. Sensitivity of the model to variations in the diets: a low-fiber diet (Diet L; 4.4% NDF) and a high-
parameters: Diets used in simulations fiber diet (Diet H; 23.8% NDF). The detailed composi-
tion of diets is in Table 4. The parameters chosen were
Diet L, % H, % set at ±50% and ±25% of their original values. The
interactions between parameter variations were not
NDF 4.4 23.8
Starch 53.8 36.3 investigated. When the degradation and absorption
Glucose 13.1 2.6 parameters were considered, values of corresponding
Proteins 20.0 26.9 parameters in SI1 and SI2 were both altered at the
Lipids 3.1 2.6 same time. Two main parts of the model were
Minerals 5.6 7.9
evaluated for their sensitivity: the first one was the
dynamics of sugars and nitrogenous fractions in the
small intestine, and the second one was the
Results metabolism in the LIC: bacterial growth and
metabolism and fecal digestibilities. The results of
Outputs of the Model. The outputs of the model are simulations with altered parameters are expressed
as follows: 1. The DM content and chemical composi- relative to the results obtained with original standard
tion of all AC at each time. 2. The chemical values, to allow a direct evaluation of the magnitude
composition of ileal and fecal flows at each time. 3. of deviations.
The flows of absorption of AA, SU, FA, VFA, AS, and The results of the sensitivity analysis on the
NN at each time. 4. The mean retention time in each metabolism of sugars in the small intestine are shown
anatomical compartment, obtained by dividing the in Table 5. The total amount of sugars absorbed in 12
mean quantity of undigestible compounds in the h was not very sensitive to the kinetics parameters:
compartment by the mean hourly dietary input of this the only important effect was a reduction of 16% for
fraction. 5. The ileal and fecal digestibility values for Diet L when the maximum velocity of glucose
all constituents, obtained by summing the flows of absorption was halved. The time and height of sugar
nutrients on a 24-h period and then dividing the result absorption peaks were more sensitive. The time of the
by the dietary input during this time. The digestibility peak was greatly affected by the maximum velocity of
of DM and energy are also calculated, by summing the absorption, especially for Diet H, which is less rich in
quantity and gross energy content, respectively, of all carbohydrates. In contrast, the fractional rate of
BSC. starch degradation and the Michaelis constant for
Sensitivity Analysis. The sensitivity of the model to absorption had little influence on the results. The
changes in selected parameters was examined for two fractional rate of gastric emptying had a more

Table 5. Sensitivity of the model to variations in the parameters: Sugars metabolisma

Time of peak Height of peak Quantity SU absorbed

Diet Variation L H L H L H
RDSTSU,SI1 and RDSTSU,SI2 −50 109 137 90 86 96 94
−25 102 114 97 95 99 98
25 98 92 102 104 101 101
50 96 86 103 106 101 102
VASU,SI1 and VASU,SI2 −50 131 144 59 78 84 94
−25 118 117 83 94 96 99
25 84 92 109 102 101 101
50 73 87 112 104 102 101
KASU,SI1 and KASU,SI2 −50 93 87 108 104 101 101
−25 97 93 103 102 101 101
25 103 106 97 98 99 99
50 106 110 94 96 99 99
RF*,STO.SI1 −50 98 110 76 71 101 100
−25 99 104 89 85 100 100
25 100 97 108 115 99 100
50 100 96 114 128 98 100
RF*,SI1.SI2 and RF*,SI2.LIC −50 122 97 104 106 103 103
−25 108 97 102 103 101 102
25 96 104 97 98 99 98
50 93 108 95 96 97 97
aOriginal parameterization = 100.

Downloaded from jas.fass.org by guest on July 14, 2011

 MODEL OF DIGESTION IN PIGS 1881
Table 6. Sensitivity of the model to variations in the parameters: Nitrogen metabolisma

Time of peak Height of peak Ileal digestibility

Date Variation L H L H L H
RDPRAA,SI1 and RDPRAA,SI2 −50 118 115 94 91 90 90
−25 106 105 96 97 96 96
25 95 96 102 102 102 102
50 93 94 104 103 104 104
VAAA,SI1 and VAAA,SI2 −50 101 93 53 53 58 55
−25 100 100 78 77 84 82
25 96 98 120 122 110 113
50 92 96 137 141 116 121
KAAA,SI1 and KAAA,SI2 −50 92 91 111 109 108 108
−25 95 96 105 104 104 104
25 102 102 96 96 97 97
50 106 105 93 93 94 94
REPR,* and REAA,* and RENN,* −50 100 100 98 99 104 105
−25 100 100 99 99 102 102
25 100 100 101 101 98 97
50 99 99 102 101 95 95
RF*,STO.SI1 −50 123 126 91 93 104 105
−25 111 112 96 97 103 104
25 90 90 103 103 96 96
50 82 83 106 105 93 92
RF*,SI1.SI2 and RF*,SI2.LIC −50 119 126 109 109 116 120
−25 105 107 104 104 107 108
25 98 95 96 96 94 93
50 95 94 93 93 90 88
aOriginal parameterization = 100.

important effect on the height than on the time of The nature of the diet used for the evaluation of the
peak absorption. The transit parameters in the small sensitivity had little effect on the results.
intestine were less influential. Validation Through Experimental Data. Firstly, the
There seemed to be a nonnegligible sensitivity to transit parameters have to be validated, because of
basic parameters of nitrogen metabolism in the small their relatively high influence on the results. As seen
intestine (Table 6). The ileal digestibility was espe- previously, this is particularly true for the gastric
cially affected by the maximum velocity of AA emptying parameter. Therefore it was checked on 18
absorption and to a lesser extent by the transit in the results not used for its original determination (Low et
small intestine. The largest effects on time and height al., 1985, Trial 2; Rainbird 1986, Diets HEM/HEMG).
of peaks of AA absorption were the velocity and the The mean fractional rate of gastric DM emptying was
transit parameters. This is a weakness of the model found to be 3.46 × 10−3 compared with 3.6 × 10−3 in the
because the parameters used for AA absorption are model. The difference ( 4 % ) is very small, considering
mean values. All AA do not behave the same, and the simulated consequences of the 25% changes of the
there are interactions because of competition for sensitivity analysis. Therefore this parameter can be
absorption. It would therefore be difficult to represent considered as reliable. Another basic verification is the
accurately the absorption of one individual AA in the validation of the quantity of matter present in the
present state of the model. The rate of endogenous different parts of the digestive tract and the flows of
secretions had little effect on the kinetics parameters matter between these compartments. The outputs of
and ileal digestibility of proteins. the model were compared with those in several
The outcome of the LIC fermentation (Table 7 ) was publications. In the experiments of Rérat and Lougnon
not sensitive to the parameters. The transit (1963), the quantities of DM in the stomach, the
parameters influenced only the mean retention time small intestine and the large intestine were followed
and the quantity of microbes in LIC. The fractional by serial slaughters of animals 1 to 8 h after ingestion
rate of fiber degradation influenced the microbial of a purified meal. The simulation of a comparable diet
metabolism and the overall digestibility of NDF. The predicts small intestine DM to be 170 g at maximum
most influential parameter was the efficiency of and 81 g after 8 h, when the reported data are 184.2 g
glucose utilization for microbial growth, which had a and 76.5 g, respectively. Similarly, the large intestine
fairly high effect on the quantity of bacteria and the DM ranged from 356 to 384 g in the simulations and
VFA production. from 265.2 to 400.2 g in the literature. However, if we

Downloaded from jas.fass.org by guest on July 14, 2011

1882 BASTIANELLI ET AL.

Table 7. Sensitivity of the model to variations in the parameters: Metabolism in LICa

QMI,LIC at Quantity VFA

MRT in LIC t = 12h absorbed FD of DM FD of CW FD of PR
Varia-
Diet tion L H L H L H L H L H L H
RF*,FEC −50 119 115 117 118 101 104 100 100 104 105 100 100
−25 108 107 107 107 101 102 100 100 101 102 100 100
25 95 96 95 94 100 99 100 100 99 97 100 100
50 90 93 90 90 99 98 100 99 98 96 100 100
KLIC −50 110 98 109 97 101 100 100 100 101 98 100 100
−25 105 100 104 98 100 100 100 100 101 99 100 100
25 96 102 96 101 100 101 100 100 99 100 100 100
50 93 103 93 103 99 101 100 100 99 100 100 100
RDCW.SU,LIC −50 97 94 92 78 95 85 100 97 80 67 100 102
−25 99 98 97 90 98 94 100 99 92 86 100 101
25 101 103 102 108 101 106 100 101 105 109 100 99
50 102 105 104 114 102 109 100 101 109 117 100 99
YG −50 114 108 59 55 117 118 102 104 102 102 104 107
−25 106 104 81 79 108 109 101 102 101 101 102 103
25 95 97 117 119 92 93 99 98 99 98 98 97
50 90 94 132 136 84 85 98 96 98 96 96 94
aFD = fecal digestibility. Original parameterization = 100.

compare the kinetics of DM passage in the ileum, the model than in the data of Giusi. This is also likely due
model predicts a peak time of approximately 4 h after to the high uptake of VFA by the gut wall. This was
the meal, which is less than the values reported in demonstrated in the work of Imoto and Namioka
Darcy et al. (1980), which were approximately 5 to 7 (1978b), who found 57% uptake of acetate by the gut
h. with a high-carbohydrate diet. After consideration of
The results of Giusi (1986) were considered for the this factor, the data seem to have a good range of
assessment of the kinetics of nutrient absorption. Her values, with a constant production of VFA showing a
experiments associate the measurement of the digesti- minimum just after the meal and an increase leading
bility of a well-defined diet to the appearance in the to peak values several hours after. Another explana-
portal blood of glucose, amino nitrogen, and VFA. The tion is that the production of VFA in the large
comparison with the simulations is shown in Figure 4. intestine is lower with purified diets such as those
A preliminary remark is that our model simulates the used by Giusi, due to a higher ileal digestibility of
quantity of nutrients that have left the lumen of the purified starch and a lower fiber digestibility than in
intestines, whereas the data of Giusi concern the actual diets (Rérat, unpublished data).
nutrients that arrive in the blood. The gut wall Comparison to Digestibility Trials. The comparison
probably delays arrival in the blood, and it may reduce with aggregative criteria allows a check to determine
the quantities entering the blood by its uptake. whether the outputs of the model are in the range in
Therefore, we can expect a difference in the absolute the literature for the classical zootechnical criteria
levels between the results. This would be a constant assessed on a daily basis. The digestibility of energy
ratio if the uptake is proportional to the quantities ( DE) of several diets differing in NDF content was
passing through the gut wall. Figure 4 suggests that assessed in the model. Figure 5 shows the comparison
the whole dynamic behavior of the model agrees of the predicted values with some classical results
rather well with the results of Giusi. For glucose (Noblet et al., 1989). The relationship between DE
absorption, the peak of glucose absorption is in both and NDF content of the diet was fairly well adjusted
cases during the second hour, but it is higher in the to a linear regression in both cases, but the slope was
simulation. The data show an overestimation in the underestimated with the model although the intercept
model only at the highest absorption levels. The ratio was approximately the same. Such data suggest that
between the experimental and simulated values is not the problem comes mainly from biases in the value of
constant, suggesting either a nonproportional uptake the potentially digestible part of total fiber and also in
by the gut wall or an overestimation of the Vmax their degradation rate, which are both constant values
parameter of glucose absorption rate. The kinetics of in the model presented. This is probably not a problem
AA absorption also had a satisfactory shape, but with of transit evaluation, because the transit times found
an overestimation by the model. The flow of VFA in the model are in the range of those of the literature
absorption was approximately two times higher in the and respond in the same way to NDF level, as seen in

Downloaded from jas.fass.org by guest on July 14, 2011

 MODEL OF DIGESTION IN PIGS 1883

Figure 5. Relationship between fecal digestibility of

energy and NDF content of the diet. (o, simulations; ÿ,
Noblet et al., 1989).

experimental data and to a deliberate simplification of

aspects known for their high level of complexity.
Validation and Results. The marginal sensitivity of
the model to individual changes in the parameters
seemed fairly low: the relative variation of the
response was generally much less than the variation
of the basic parameters. However, some parameters
seem to have a non-negligible effect. This was in
particular the case of the maximum velocities of
absorption. One general comment that can be made is
that the parameters that were more difficult to
estimate from the literature have a limited effect on
the response: this is, for example, the case of the Km
Figure 4. Kinetics of absorption of a) glucose, b) for AA absorption. Some values can be estimated for
amino acids, and c) VFA in the model compared with individual AA (Adibi, 1969), but the relation is
experimental results for similar diets. Black bars, Giusi theoretical for the pool of AA. The consequences of the
(1986); open bars, results of the simulations. poor knowledge available on this parameter are
limited because the sensitivity analysis showed that
the effects are five times less than the change in
parameter. The situation is the same for the transit
Figure 6, which shows a number of literature results parameters in LIC, which means that the approxima-
about mean retention time and the trend of the model tion generated by the choice of a conceptual function
response when tested over a wide range of diets. in that case has probably no dramatic consequences on
the validity of the model. The maximum velocity of
absorption flow is much more sensitive but better
Discussion known from the literature. This is also the case for the
degradation parameters, which are mean values in the
The current model constitutes a first attempt to model. However, the known variations from one feed
simulate mechanistically the digestion in pigs. It is to another in the susceptibility to enzymatic degrada-
therefore an interesting complementary step to meta- tions can have a rather important effect and should
bolic models such as that of Pettigrew et al. therefore be included as input to the simulations for
(1992a,b), which runs at small time steps but an improved accuracy.
considers the dietary input as constant through the The external validation presented here is limited to
day. In its present state, our model provides a some aspects of general behavior of the model. Data
comprehensive framework for further inclusion of are too scarce in the literature to allow a relevant
knowledge. It already yields fairly sound prediction for evaluation of the quality of the results of the model on
some digestibility coefficients, transit time, and pat- a dynamic point of view. Most data published on
tern of nutrient absorption. However, it has several absorption of nutrients from the intestine of pigs or
main limitations due in particular to a certain lack of humans concern very particular cases (e.g., Mahé et

Downloaded from jas.fass.org by guest on July 14, 2011

1884 BASTIANELLI ET AL.

Therefore all phenomena were considered to depend

only on DM.
The choice of a compartmental representation
seemed quite evident in the case of the stomach, which
is an actual mixing compartment, with an emptying
pattern that is satisfactorily described by a first-order
process when the feed consists in a single phase of
small particles. When more than one phase is present
(feed with pieces), the hypothesis of a single DM
compartment in the stomach is less sustainable, and
two (or more) stomach DM compartments have to be
considered, with possible passage from one to another
through reduction of particle size and solubilization.
This kind of approach has been proposed for modeling
Figure 6. Relationship between mean retention time the human stomach by Bernier et al. (1988).
(MRT, hours) and NDF content of the diet. (−o−), (——), Events in the small intestine are modeled with
prediction of the model. Data from Canguilhem and more difficulty. In fact, in this part of the digestive
Labie (1977), Castle and Castle (1956, 1957), Cherbut et tract, digesta are propulsed by peristaltic movements,
al. (1988), Ehle et al. (1982), Fioramonti and Bueno which also have a local mixing effect. One solution to
(1980), Furuya and Takahashi (1975), Furuya et al. model it would be to consider a continuous progression
(1978), Kuan et al. (1983), Pond et al. (1986), Potkins et of digesta along the small intestine, with introduction
al. (1991), Roth and Kirchgessner (1985), Sandoval et al. of diffusion processes. Such an approach has already
(1987), and Stanogias and Pearce (1985). been suggested by France et al. (1993) for the study
of the kinetics of appearance of unabsorbable markers
in feces. In principle, this approach would probably
al., 1992) and very seldom reflect a realistic situation allow a more mechanistic consideration of the delaying
of nutrient absorption after a real standard meal. and smoothing effects (viscosity, etc.), but in the case
Even when this latter kind of data is available (e.g., of the system defined here, it would lead to a more
Rérat et al., 1988), most of the time it concerns the complicated formalization. Moreover, even if the
enrichment of portal blood in the specified nutrient structure of the model seems closer to reality, some
rather than its departure from the lumen of the gut, underlying hypotheses are in fact not valid in the case
which is the variable simulated in the current model. of pig’s small intestine, e.g., consideration of a perfect
On the other hand, there is a lot of published data nondeformable cylinder with no propulsion
about nutrient digestibilities to which the model can phenomena, homogeneous medium, and laminar flow
with no mixing. In addition, some parameters seem to
be compared, but they are aggregative criteria that
be poorly documented (e.g., viscosity parameters of
cannot validate the dynamics of the model’s outputs,
digesta, diffusion parameters of all constituents).
and therefore we limited this comparison to digestibil-
Other reasons that limit the use of the diffusion
ity of energy as a sole example in this article.
models in the present case are the facts that the
Structure of the Model. The total number of
quantity of moving matter is not conserved along the
compartments is 44. However, the hierarchical struc-
small intestine and that there is a relation between
ture of the model, which consists of four DM anatomi-
the rate of transit and absorption, both linked to the
cal compartments with description of the chemical
quantity of matter.
composition of the DM, makes it much less compli- Another possible way of modeling transit is to
cated than a model that would include 44 independent divide the small intestine into a set of numerous small
compartments. In that sense, the model can be compartments with an assumption of a mass action
considered as a simple one. law for emptying. But, apart from the complexity of
The choice of considering only the DM content was such a model, a main objection to this approach is that
made to respect the aim of model simplicity. The the transit in each portion also depends on the
water movements depend on the osmolarity at each quantity of matter in the adjoining portions, leading to
level of the digestive tract, and therefore on the ion complicated and insufficiently documented interac-
and mineral transactions, on the physico-chemical tions. Lastly, the degradation and absorption
properties of the digesta (water-holding capacities), parameters are not known in the different portions. In
on some parameters related to the metabolism of the consequence, the current choice of a limited number of
animal, on the feed and drinking water, and other mixing compartments seemed to be the most reasona-
factors. In addition to complicating the model, these ble one for our purposes, even if it has not a precise
factors are very hazardous to introduce in the model, biological appearance.
because this information is very seldom available in The use of a delay function for the emptying of the
the experiments used to build or validate the model. intestinal compartment accounts for the time of

Downloaded from jas.fass.org by guest on July 14, 2011

 MODEL OF DIGESTION IN PIGS 1885
transit through this part of the digestive tract. been published, a comprehensive study of the sig-
Literature data were used to provide average reten- nificance and accuracy of these tests for our purpose
tion time values; however, some aspects linked to the has to be performed.
order of the delay are still in question. The aim of However, the in vitro approach is still insufficient,
simplicity of the model also led us to voluntarily and because one aim of the model is to provide an
ignore some major regulatory factors of transit and estimation of the kinetics of nutrient absorption (e.g.,
digestion. In a model aimed at being more mechanistic for use in a metabolic model), one solution would be to
and more accurate, some of them would have to be take as input for the simulation the ileal digestibilities
taken into account. In particular, because we stressed of the different fractions of the feed and to use these
the importance of the gastric emptying on the kinetics values to adjust the degradation parameters. The
of absorption, the role of such metabolites such as fat prediction of the nutrient input pattern for the
or AA on the rate of gastric emptying is important. metabolism of the animal also requires the model to
The “delay” function used here has the advantage of account for the uptake of nutrients by the gut wall.
delaying and smoothing the transit function, without Several authors have stressed the importance of this
introducing additional compartments. Because the organ in the energy expenditure and also in the
parameterization would not be supported by ex- protein turnover in the organism (e.g., Sève et al.,
perimental data if such compartments were in- 1993). This means that a large part of the nutrients
troduced, the empirical character of the delay function disappearing from the lumen is used by the gut wall to
does not result in a decreased accuracy. However, the sustain its intense metabolic activity. Another specific
shape of the DM passage at the ileum is fairly close to role of the gut wall that requires more data is the
those observed in the literature (e.g., Giusi, 1986) delay that it creates between nutrient disappearance
and can be considered acceptable in a first step. from the lumen and appearance in the portal blood.
Endogenous Secretions. Much literature exists on This latter aspect should be evaluated for a prediction
aspects related to endogenous secretions. However, of the real availability of nutrients for metabolism.
the data are very poorly integrated into a consistent This would allow a direct validation of the output by
quantitative knowledge. Though some coherent infor- data of nutrient appearance in the blood.
mation is becoming available for nitrogen (review of One problem that arises from the compartmental
Souffrant, 1991), this is not the case for other structure based on DM is that the quantities and
compounds (Juste, 1982). The choice here to index concentrations of the nutrients are not representative
endogenous secretions primarily on the passage of DM of the real ones in each place of the small intestine.
is coherent with the available literature: it accounts This is a problem because the data on absorption of
nutrients in the literature very often relate to luminal
for the linear relationship between endogenous secre-
concentrations of nutrients. The problem when con-
tions and DM intake and also takes into account the
sidering concentrations is that in each little portion of
higher level of endogenous secretions when the diet is
the small intestine, the absorption is basically linked
less digestible (e.g., for higher fiber content).
to the concentration itself. The concentration that can
However, some other characteristics of the diet could
be calculated in a global compartment is not the mean
also be considered if quantitative data are made
of the local concentrations. It cannot therefore be used
available: it is the case of the other chemical
to represent phenomena linked with these concentra-
compounds of the diet (lipids, proteins,...) and an-
tions. Furthermore, because the water is not taken
tinutritional factors (Schulze, 1994).
into account, the “concentrations” that could be
Degradation and Absorption Processes. The current calculated are only proportions of DM and can hardly
model was conceived to be able to simulate the main be considered as the driving forces for absorption
laws of response of digestion to diet variations. Fairly processes. This is why the choice of considering
satisfactory general data were obtained for these quantities of matter rather than concentrations was
aspects. However, more accuracy is obviously needed made. Even if they can be regarded as less relevant at
to connect feed analytical data to parameters of the a local level, they have the advantage of being additive
model allowing one to obtain good estimations for and can consequently be used in global aggregative
their feed value. For these aspects, this model compartments.
presents in its present state less accuracy than some
available empirical models of prediction of digestibility
values. For instance, the degradation rates of the feed Implications
constituents are determining parameters of the feed
value. It seems therefore important to use the results With an aim of integrating knowledge on the
of some in vitro tests of digestibility of the main metabolism of animals into operational or (in a first
constituents as input parameters for the model, in step) research models, the possibility to take into
order to better predict both the digestibility and the account the dynamics of entry of the nutrients in
kinetics of nutrient absorption. However, because no metabolism is essential. To date, much data concern-
completely satisfactory method for in vitro tests has ing nutrient absorption in pigs, rats, and humans has

Downloaded from jas.fass.org by guest on July 14, 2011

1886 BASTIANELLI ET AL.

been produced but not integrated into a coherent Darcy, B., J. P. Laplace, and P. A. Villiers. 1980. Digestion dans
framework. The current model constitutes a first step l’intestin grêle chez le porc. 2- Cinétique comparée de passage
des digesta selon le mode de fistulation, iléocaecale ou iléo-
toward new feed units expressed in terms of flows of colique post valvulaire, dans diverses conditions d’alimentation.
absorbed nutrients. The flow diagram of the model has Ann. Zootech. 29:147.
been established for pigs, but it could as well be Darragh, A. J., P. D. Cranwell, and P. J. Moughan. 1994. Absorption
applied to other monogastric animals such as poultry, of lysine and methionine from the proximal colon of the piglet.
horses, and humans with different parameterization Br. J. Nutr. 71:739.
Dierick, N., I. Vervaeke, J. Decuypere, and H. K. Henderickx. 1990.
but minor changes in its structure.
Bacterial protein synthesis in relation to organic matter diges-
tion in the hindgut of growing pigs. Contribution of hindgut
fermentation to the total energy supply and growth perfor-
Literature Cited mances. J. Anim. Physiol. Anim. Nutr. 63:220.
Dijkstra, J., and J. France. 1995. Modeling and methodology in
Adibi, S. A. 1969. The influence of molecular structure of neutral animal science. In: A. Danfaer and P. Lescoat (ed.). Proceed-
amino acids on their absorption kinetics in the jejunum and ings of the IVth International Workshop on Modeling nutrient
ileum of human intestine in vivo. Gastroenterology 56:903. utilisation in farm animals. p 9. NIAS, Foulum, Denmark.
Argenzio, R. A. 1982. Volatile fatty acid production and absorption Ehle, F. R., J. L. Jeraci, J. B. Robertson, and P. J. Van Soest. 1982.
from the large intestine of the pig. In: J. P. Laplace, T. Corring, The influence of dietary fiber on digestibility, rate of passage
and A. Rérat (Ed.) Digestive Physiology in the Pig. p 207. and gastrointestinal fermentation in pigs. J. Anim. Sci. 55:
INRA, Versailles, France. 1071.
Auffray, P., J. Martinet, and A. Rérat. 1967. Quelques aspects du Fioramonti, J., and L. Bueno. 1980. Motor activity in the large
transit gastro-intestinal chez le porc. Ann. Biol. Anim. Biochim. intestine of the pig related to dietary fibre and retention time.
Biophys. 7:261. Br. J. Nutr. 43:155.
Bach Knudsen, K. E., B. Borg Jensen, J. O. Andersen, and I. France, J., J.H.M. Thornley, R. C. Siddons, and M. S. Dhanoa. 1993.
Hansen. 1991. Gastrointestinal implications in pigs of wheat On incorporating diffusion and viscosity concepts into compart-
and oat fractions. 2- Microbial activity in the gastrointestinal mental models for analysing faecal marker excretion patterns
tract. Br. J. Nutr. 65:233. in ruminants. Br. J. Nutr. 70:369.
Bastianelli, D., D. Sauvant, and A. Rérat. 1994. Bases d’un modèle Friedman, H. I., and B. Nylund. 1980. Intestinal fat digestion,
mécaniste simple de la digestion chez le porc. Journ. Rech. absorption and transport. Am. J. Clin. Nutr. 33:1108.
Porcine Fr. 26:227. Furuya, S., K. Sakamoto, T. Asano, S. Takahashi, and K. Kameoka.
Bernier, J. J., J. Adrian, and N. Vidon (ed.) 1988. Les aliments dans 1978. Effects of added dietary sodium polyacrylate on passage
le tube digestif. Editions Doin, Paris, France. rate of markers and apparent digestibility by growing swine. J.
Black, J. L., R. G. Campbell, I. H. Williams, K. J. James, and G. T. Anim. Sci. 47:159.
Davies. 1986. Simulation of energy and amino acid utilisation Furuya, S., and S. Takahashi. 1975. Rate of passage of chromic
in the pig. Res. Dev. Agric. 3:121. oxyde and polyethylene Glycol and digestibility in the digestive
Canguilhem, R., and C. Labie. 1977. Variation de la durée du transit tract of pigs. Jpn. J. Zootech. Sci. 46:630.
intestinal chez le porc en fonction de la teneur de la ration en Gaudichon, C. 1994. Marquage à l’azote 15N des fractions azotées
cellulose. Rev. Med. Vet. 12:1669. endogène et exogène dans la lumière intestinale chez l’homme
Castle, E. J., and M. E. Castle. 1956. The rate of passage of food et le porc miniature: application à la digestion des proteines de
through the alimentary tract of pigs. J. Agric. Sci. 47:196. produits laitiers ( 15N labeling of endogenous and exogenous
Castle, E. J., and M. E. Castle. 1957. Further studies of the rate of nitrogenous fractions in the intestinal lumen of man and minia-
passage of food through the alimentary tract of pigs. J. Agric. ture pig: Application to digestion of dairy products proteins).
Sci. 48:106. PhD Dissertation. INAPG, Paris, France.
Cherbut, C., J. L. Barry, M. Wyers, and J. Delort-Laval. 1988. Effect Gill, M., D. E. Beever, and J. France. 1989. Biochemical bases
of the nature of dietary fibre on transit time and faecal excre- needed for the mathematical representation of whole animal
tion in the growing pig. Anim. Feed Sci. Technol. 20:327.
metabolism. Nutr. Res. Rev. 2:181.
Clemens, E. T., C. E. Stevens, and M. Southworth. 1975. Sites of
Giusi, A. 1986. Influence du niveau alimentaire et de la composition
organic acid production and pattern of digesta movement in the
du régime sur la production intestinale et l’absorption des
gastrointestinal tract of swine. J. Nutr. 105:759.
acides gras volatils chez le porc éveillé. (Influence of the level
Clément, J. 1980. Intestinal absorption of triglycerols. Reprod. Nutr.
of feeding and of the composition of feed on the production and
Dev. 20:285.
absorption of volatile fatty acids in the conscious pig). PhD
Corring, T., A. Aumaitre, and A. Rérat. 1972. Fistulation per-
Dissertation. University of Paris VI, Paris, France.
manente du pancréas exocrine chez le porc. Application:
Hungate, R. E. 1966. The Rumen and Its Microbes. Academic Press,
Réponse de la sécrétion pancréatique au repas. Ann. Biol.
New York.
Anim. Biochim. Biophys. 12:109.
Cuber, J. C., C. Kabore, and J. P. Laplace. 1981. Evacuation gast- Imoto, S., and S. Namioka. 1978a. VFA metabolism in the pig. J.
rique comparés de l’orge et du blé broyés chez le porc. Reprod. Anim. Sci. 47:479.
Nutr. Dev. 21:773. Imoto, S., and S. Namioka. 1978b. VFA production in the pig large
Cuber, J. C., and J. P. Laplace. 1979. Evacuation gastrique de la intestine. J. Anim. Sci. 47:467.
matière sèche d’un régime semi-purifié à base d’amidon de maı̈s Juste, C. 1982. Apports endogènes par les secrétions digestives chez
chez le porc. Ann. Biol. Anim. Bioch. Biophys. 19:899. le porc. In: J. P. Laplace, T. Corring, and A. Rérat (ed.) Diges-
Cuber, J. C., J. P. Laplace, and P. A. Villiers. 1980. Fistulation de tive Physiology in the Pig. p 155. INRA, Versailles, France.
l’estomac et contenus gastriques résiduels après ingestion d’un Keys, J. E., Jr., and J. V. DeBarthe. 1974a. Cellulose and hemicellu-
régime semi-purifié à base d’amidon de maı̈s chez le porc. lose digestibility in the stomach, small intestine and large
Reprod. Nutr. Dev. 20:1161. intestine of swine. J. Anim. Sci. 39:53.
Czerkawski, J. W. 1976. Chemical composition of microbial matter Keys, J. E., Jr., and J. V. DeBarthe. 1974b. Site and extent of
in the rumen. J. Sci. Food Agric. 27:621. carbohydrate, dry matter, energy and protein digestion and the
Darcy, B. 1984. Availability of amino acids in monogastric animals. rate of passage of grain diets in swine. J. Anim. Sci. 39:57.
Variations of digestive origin. Diabete Metabolisme (Paris) 10: Kuan, K. K., G. Stanogias, and A. C. Dunkin. 1983. The effect of
121. proportion of cell wall material from lucerne leaf on apparent

Downloaded from jas.fass.org by guest on July 14, 2011

 MODEL OF DIGESTION IN PIGS 1887
digestibility, rate of passage and gut characteristics in pigs. Rérat, A. 1981a. Digestion and absorption of nutrients in the pig.
Anim. Prod. 36:201. World Rev. Nutr. Diet. 37:229.
Laplace, J. P. 1970. Mensurations in vivo et post mortem et crois- Rérat, A. 1981b. Contribution du gros intestin à la digestion des
sance de l’intestin grêle chez le porc. Ann. Zootech. 19:465. glucides et des matières azotées chez le monogastrique omni-
Laplace, J. P., and. J. C. Cuber. 1984. Déafférentation vagale totale vore. Reprod. Nutr. Dev. 21:815.
et évacuation gastrique chez le porc. Reprod. Nutr. Dev. 24:655. Rérat, A., and J. Lougnon. 1963. Etudes sur le transit digestif chez
Laplace J. P., O. Pons, J. C. Cuber, and C. Kaboré. 1985. Evacuation le porc. Ann. Biol. Anim. Biochim. Biophys. 3:21.
gastrique de l’orge et du blé chez le porc. Ann. Zootech. (Paris). Rérat, A., P. Vaissade, and P. Vaugelade. 1988. Absorption kinetics
34:265. of dietary hydrolysis products in conscious pigs given diets with
Low, A. G. 1990. Nutritional regulation of gastric secretion, diges- different amounts of fish proteins. 1. Amino nitrogen and glu-
tion and emptying. Nutr. Res. Rev. 3:229. cose. Br. J. Nutr. 60:91.
Low, A. G., R. J. Pittman, and R. J. Elliott. 1985. Gastric emptying Rivest, J., C. Pomar, and J. F. Bernier. 1994. Simulation of digestion
of barley-soyabean diets in the pig: Effects of feeding level, of proteins in the small intestine of growing-finishing pigs. J.
supplementary maize oil, sucrose or cellulose, and water in- Anim. Sci. 72(Suppl.1):264 (Abstr.).
take. Br. J. Nutr. 54:437. Roth, F. X., and M. Kirchgessner. 1985. Verdaulichkeit und intes-
Low, A. G., and T. Zebrowska. 1989. Digestion in pigs. In: H. D. tinale Passagerate beim Schwein in Abhängigkeit vom Füt-
Bock, B. O. Eggum, A. G. Low, O. Simon, and T. Zebrowska terungsniveau und Rohfasergehalt des Futters. Z. Tierphysiol.
(ed.) Protein Metabolism in Farm Animals. Oxford University Tierernaehr. Futtermittelkd. 53:254.
Press, Oxford, U.K. Sambrook, I. E. 1979a. Studies on digestion and absorption in the
Mahé, S., J. F. Huneau, P. Marteau, F. Thuillier, and D. Tomé. 1992. intestines of growing pigs. 8. Measurements of the flow of total
Gastro-ileal nitrogen and electrolyte movements after bovine lipid, ADF and volatile fatty acids. Br. J. Nutr. 42:279.
milk ingestion in humans. Am. J. Clin. Nutr. 56:410. Sambrook, I. E. 1979b. Studies on digestion and absorption in the
Mason, V. C. 1984. Metabolism of nitrogenous compounds in the intestines of growing pigs. 7. Measurements of the flow of total
large gut. Proc. Nutr. Soc. 43:45. carbohydrate, total reducing substances and glucose. Br. J.
Matthews, D. M. 1991. Protein Absorption: Development and Nutr. 42:267.
Present State of the Subject. Wiley-Liss, New York. Sandoval, R. A., T. K. Nielsen, and P. H. Sorensen. 1987. Effects of
Merry, R. J., and A. B. McAllan. 1983. The chemical composition of fibre on nutrient digestion and time of passage in growing pigs.
mixed bacteria separated from different phases of rumen Acta Agric. Scand. 37:367.
digesta. Proc. Nutr. Soc. 42:49A (Abstr.). Sauvant, D. 1992. La modélisation systémique en nutrition. Reprod.
Miller, L. J., J. R. Malagelada, and L. W. Go. 1978. Postprandial Nutr. Dev. 32:217.
duodenal function in man. Gut. 19:699. Schulz, A. R. 1978. Simulation of energy metabolism in the simple-
Moran, E. T. (Ed.) 1982. Comparative Nutrition of Fowl and Swine. stomached animal. Br. J. Nutr. 39:235.
The Gastrointestinal Systems. University of Guelph, Guelph, Schulze, H. 1994. Endogenous ileal nitrogen losses in the pig:
Ontario, Canada. Dietary factors. Ph.D. dissertation. University of Wageningen,
Moughan, P. J., and W. C. Smith. 1984. Prediction of dietary protein The Netherlands.
quality based on a model of the digestion and metabolism of Sève, B., O. Ballèvre, P. Granier, J. Noblet, J. Prugnaud, and C.
nitrogen in the growing pig. N.Z. J. Agric. Res. 27:501. Obled. 1993. Recombinant porcine somatotropin and dietary
Nixon, S. E., and G. E. Mawer. 1970. The digestion and absorption of protein enhance protein synthesis in growing pigs. J. Nutr. 123:
protein in man. Br. J. Nutr. 24:227. 529.
Noblet, J., H. Fortune, S. Dubois, and Y. Henry. 1989. Nouvelles Souffrant, W. B. 1991. Endogenous nitrogen losses during digestion
bases d’estimation des teneurs en énergie digestible, métabolis- in pigs. In: Proceedings of the Vth International Symposium on
able et nette des aliments pour le porc. INRA éditions, Paris, Digestive Physiology in the Pig. EAAP publ. No. 54, p 147.
France. Pudoc, Wageningen, The Netherlands.
Pettigrew, J. E., M. Gill, J. France, and W. H. Close. 1992a. A Stanogias, G., and G. R. Pearce. 1985. The digestion of fibre by pigs.
mathematical integration of energy and amino acid metabolism 1- The effects of amount and type of fibre on apparent digesti-
of lactating sows. J. Anim. Sci. 70:3742. bility, nitrogen balance and rate of passage. Br. J. Nutr. 53:513.
Pettigrew, J. E., M. Gill, J. France, and W. H. Close. 1992b. Evalua- Storm, E., and E. R. Orskov. 1983. The nutritive value of rumen
tion of a mathematical model of lactating sow metabolism. J. micro-organisms in ruminants. I- Large scale isolation and
Anim. Sci. 70:3762. chemical composition of rumen micro-organisms. Br. J. Nutr.
Pomar, C., D. L. Harris, and F. Minvielle. 1991a. Computer simula- 50:463.
tion model of swine production systems: I. Modeling the growth Tomassone, R., and J. P. Laplace. 1973. Essai d’utilisation des
of young pigs. J. Anim. Sci. 69:1468. systèmes compartimentés dans une étude sur la physiologie de
Pomar, C., D. L. Harris, and F. Minvielle. 1991b. Computer simula- la digestion. Ann. Biol. Anim. Biochim. Biophys. 13:747.
tion model of swine production systems: II. Modeling body Vervaeke, I. J., N. A. Dierick, D. I. Demeyer, and J. A. Decuypere.
composition and weight of female pigs, fetal development, milk 1989. Approach to the energetic importance of fibre digestion in
production, and growth of suckling pigs. J. Anim. Sci. 69:1489. pigs. II- An experimental approach to hindgut digestion. Anim.
Pond, W. G., K. R. Pond, W. C. Ellis, and J. H. Matis. 1986. Markers Feed Sci. Technol. 23:169.
for estimating digesta flow in pigs and the effects of dietary Vodovar, N., J. Flanzy , and C. François. 1964. Intestin grèle du
fiber. J. Anim. Sci. 63:1140. porc. Ann. Biol. Anim. Biochim. Biophys. 4:27.
Potkins, Z. V., T.L.J. Lawrence, and J. R. Thomlinson. 1991. Effects Warner, A.C.I. 1981. Rate of passage of digesta through the gut of
of structural and non-structural polysaccharides in the diet of mammals and birds. Nutr. Abstr. Rev. Series B 51:789.
the growing pig on gastric emptying rate and rate of passage of Whittemore, C. T., and R. H. Fawcett. 1974. Model responses of the
digesta to the terminal ileum and through the total gastrointes- growing pig to the dietary intake of energy and protein. Anim.
tinal tract. Br. J. Nutr. 65:391. Prod. 19:221.
Pugh-Roberts. 1986. Professional Dynamo plus. Reference manual. Whittemore, C. T., and R. H. Fawcett. 1976. Theoretical aspects of a
Pugh-Roberts Associates, Cambridge, MA. flexible model to simulate protein and lipid growth in pigs.
Rainbird, A. L. 1986. Effect of guar gum on gastric emptying of test Anim. Prod. 22:87.
meals of varying energy content in growing pigs. Br. J. Nutr. Zebrowska, T., and F. Horszczaruk. 1975. Wplyw jedno- i
55:99. dwukrotnego karmienia na ilosc i sklad tresci jelita cienkiego
Rainbird, A. L., and A. G. Low. 1986. Effect of various types of swin (Effect of feeding frequency on the amount and composi-
dietary fibre on gastric emptying in growing pigs. Br. J. Nutr. tion of digesta in the small intestine of pigs). Rocz. Nauk Roln.
55:111. Ser. B 96:91.

Downloaded from jas.fass.org by guest on July 14, 2011

Citations This article has been cited by 7 HighWire-hosted articles:
http://jas.fass.org/content/74/8/1873#otherarticles

Downloaded from jas.fass.org by guest on July 14, 2011