R.G. LENTLE
Institute of Nutrition and Human Health,
Massey University, Palmerston North,
New Zealand
This article relates the physical properties of feeds and digesta to digestive efficiency in poultry and outline several works on the effect of the physical properties of digesta on flow. Digesta, which contains significant proportions of solid matter, has high non Newtonian viscosity which suppresses turbulent mixing and causes laminar flow to prevail; a situation which runs contrary to the predictions of chemical reactor theory. Digesta that contains significant proportions of flexible fibrous material behaves as a weak gel that allows extrusion of the liquid phase during peristalsis. This augments plug flow by disengagement of adhesion of the digesta plug to mucus and mucosa but also may aid permeation of the plug by digestive secretions. Preliminary work shows that efficient digestion of digesta that contains significant quantities of rigid solid material is related to the ease of passive permeation which is governed by particle size and by the viscosity of the fluid phase.
The gross effects of the physical properties of feeds on digestibility and disease
A species survival though evolutionary time is related to its ability to gain nutrients efficiently. A number of mechanisms may serve to optimise nutrient gain, for example the search for food items, the choice between food items, and the nutrient composition of food items. In nature these mechanisms are ameliorated by satiety thus giving time for other activities necessary for survival. The exigencies of commercial production, in selecting for nutrient gain, may reduce the efficiency of these ameliorative mechanisms. In the latter scenario, the digestive process is likely to be the ultimate arbiter of nutrient gains. Whilst nutrient gain is influenced by nutritional density, it has been increasingly recognised that the influence of particular food items on the physical properties of digesta may also modulate digestive efficiency (Hetland and Svihaus, 2001).
With regards to nutrient gain in poultry, two particular findings indicate a likely influence of the physical properties of digesta on digestibility; the fact that digestibility is adversely influenced by fibre content of feeds (Dusel et al., 1997; Malathi et al., 1997; Masonnier et al., 2001) and the fact that these adverse effects may be ameliorated by the addition of specific enzymes that degrade such fibres (Malathi and Dvegowda, 2001; Svihus and Guillard, 2002).
Other investigations have more directly correlated gross physical properties with digestibility. Thus digestibility has been related to the viscosity of the feed (Malathi and Devegowda, 2001) and to particle size (Carre et al., 1998). Similarly, a number of particular dietary elements have been identified that adversely influence viscosity and digestibility. Thus, whilst the incorporation of linseed into the diet markedly increases the viscosity of ileal digesta, decreases gross fat digestibility, digestibility of single fatty acids and dietary metabolisable energy and depresses weight gain, these adverse effects can be partially overcome by using ‘demucilaged' linseed (Alzueta et al., 2003). Again, the feeding of linseed mucilage to broiler chicks significantly increases the viscosity of their jejunal digesta and has significant adverse effects on crude fat and fatty acid digestibility and AME (Rebole et al., 2002).
Other investigations have shown that some agents that influence the physical properties of digesta may selectively impair the digestibility of particular nutrients. Thus, the addition of carboxymethyl cellulose (CMC) to the diet of growing chickens significantly increases the viscosity of the liquid phase of their digesta and the ratio of liquids to solids, and adversely effects the digestibilities of C16:0 and C18:0 fatty acids but has no effect on the digestibilities of C18:1 , C18:2 or C12:0 or C14:0 fatty acids (Smits et al., 2001). Finally it is of note that the addition of particular agents which effect the physical properties of digesta may influence pathogenesis. Thus the addition of CMC to a wheat based diet significantly reduces feed conversion efficiency and weight gain in broiler chickens infected on day 21 with Eimeria acervulina (Banfield et al., 2002).
Whilst a body of work links the physical properties of food or digesta to digestive efficiency there have been no attempts to gain an understanding of the mechanisms by which this is brought about. The following describes our investigations of some physical parameters that may influence the on flow of digesta and the efficiency of the digestive process.
The importance of mixing within the gut lumen
The advent of confocal microscopic and genetic techniques has afforded us a detailed knowledge of the mechanisms by which nutrients are transported from the intestinal lumen into the enterocyte and beyond. However, relatively little is known of the mixing environment within the intestinal lumen.
Glycocalyceal extensions of the enterocyte microvilli may enclose an ‘unstirred water layer' that necessitates the transit of nutrient and secreted substances by simple diffusion (Johnson and Gee, 1981). Depletion of the rate of nutrient entry into this layer will impair concentration gradients, extend the mean diffusion times and impair absorption efficiency. Efficient intraluminal mixing is therefore important for the maintenance of optimum rates of diffusion across the unstirred water layer and thus for the efficiency of digestion.
Apart from its influence in optimising rates of nutrient absorption, the efficiency of intralumenal mixing influences the establishment and maintenance of commensal facultative and obligate anaerobic populations of intestinal microflora (Borriello, 2002). The establishment and maintenance of a balanced ecosystem of mucosal and luminal microbial populations is increasingly recognised as important for the maintenance of appropriate immune response, the prevention of invasion by pathogens and the proper development of the juvenile gut so as to create ‘colonial resistance'(Snel et al., 2002). Moreover, a work by Jeffery Gordon of the Washington School of Medicine suggests that certain enteral microflora can boost nutrient gain from digesta and up regulate the hosts storage mechanisms by suppressing fasting induced adipocyte factor.
The early human neonatal lumen is rich in oxygen and the zone of relative anoxia in the central lumen is only established following successful enteral colonisation by facultative anaerobes (Mackie et al., 1999). In poultry there is similar early colonisation by facultative with subsequent colonisation by obligate anaerobic species (Snel, 2002). Concomitant with enteral colonisation the balance of the immune response shifts from predominantly humoral to cellular immunity (Goddeeris et al., 2002).
Changes in the viscosity of digesta may influence the ease of establishment of pathogenic microflora by influencing the ease of mixing of the marginal oxygenated and the central lumenal anoxic zones. Thus dietary supplementation with CMC increases the viscosity of ileal digesta, which may aid the establishment of enterotoxigenic E coli and Brachyspira pilocoli (Hopwood et al., 2002) in young pigs.
Theory: the physics of intraluminal mixing
Mixing of reactants within the gut can occur in two broad ways, macromolecular mixing, where masses of reactants move in relation to one another, for example by convection or turbulence, or micromolecular mixing where individual molecules of reactant admix via diffusion (Levenspiel, 1972). The latter process is slow over large distances (Crank, 1975). Thus, at body temperature, the distance moved by a diffusing molecule of sucrose in 10 hours is 0.43cm (France et al., 1993). Given that the diameter the small intestine of the chicken is of the order of one cm, then a molecule of sugar situated at the centre of the lumen would take ten hours to diffuse to the periphery for absorption.
Highly efficient macromolecular mixing can occur during turbulent flow when groups of molecules gain sufficient momentum to admix randomly with adjacent material. The ease with which this can occur is proportional to the velocity of flow and the density of material and inversely proportional to the viscosity of the medium (which latter is equivalent to the frictional force impairing the build up of momentum). The relationship between these parameters can be expressed as a Reynolds number, lower values of this number indicating a lower likelihood of turbulent flow. Watery fluids such as blood flowing in biological tubular systems tend to have very low Reynolds numbers indicating that turbulent flow is unlikely and laminar flow, where fluid layers slide over each other in an orderly manner, is most likely. Given that digesta contains significant quantities of solid material which further increases viscosity, then Reynolds numbers are likely to be very low and turbulent flow unlikely in tubular portions of the gut (Lentle et al., 2002). Thus laminar flow conditions have generally been assumed to prevail in regards to the flow of digesta through the GI tract which would effectively preclude mixing across the radial dimension. The validity of this assumption is supported by studies showing that the laminar flow model is able to accurately predict the uptake of substances from watery solutions perfused though isolated gut segments (Levitt et al., 1988).
The technique of spatiotemporal video mapping (Hennig et al., 1999) has been used to measure the extent and velocity of radial and longitudinal movements during rapid peristaltic contractions induced by lumen distension and by instillation of nutrients (Decanoic acid) (Gwynne et al., 2004). The velocity vectors obtained by these workers are generally insufficient to give Reynolds numbers associated with turbulent mixing. However, it is worthy of note that a modelling of the flow of simple watery fluids though the rat ileum using spatiotemporal data indicate that vorticeal flow may be induced by sudden reversals of flow as the contracting segment propagates along the intestine (Jeffery et al., 2003). Thus turbulent mixing may occur when the intestine is occupied by simple watery solutions.
Thus far we have considered macromolecular mixing by movement of masses of digesta in relation to one another. However, macromolecular mixing can also occur by movement of the fluid phase relative to the solid phase i.e. permeation. Such relative movement is dependent on both the physical properties of the digesta and the hydraulic conditions within the lumen. Thus, provided that the digesta will sustain compression during the application of force and not ‘flow away', compression of the matrix, (either directly by muscular contraction or indirectly via the application of hydrostatic pressure), may cause fluid phase to travel to an area of lower pressure (Weems, 1982). Spatiotemporal analysis of small intestinal contractions shows that certain patterns of small intestinal movement may be conducive to such relative movement (Gwynne et al., 2004). Thus zones of relative dilatation of the gut lumen are often found adjacent to regions in which there are stationary or propagating contractions (Gwynne et al., 2004).
Determining the conditions for optimum digestive efficiency: the use of chemical reactor models
In the latter few decades, a number of studies have been made of the manner in which the flow of reactants through the digestive tract approximate that of ideal chemical reactors under the assumption that natural selection will tend to optimise the digestive process (Penry and Jumars, 1987). Thus the rumen has been characterised as a ‘continuous stirred flow reactor' where highly efficient mixing must occur at all points within the lumen ‘global mixing'. Conversely, the small intestine should approximate a ‘plug flow reactor' (PFR) in which there is perfect radial mixing within consecutive radial ‘slices' but minimal mixing in the direction of travel i.e. axially. However, under laminar flow conditions, digestion in the small intestine will not approach such optimality as efficient mixing of the digesta within successive radial ‘slices' does not take place and there is axial mixing. Thus, those physical characteristics of digesta which promote laminar flow and decrease turbulence in successive segments of small intestine may be expected to decrease digestive efficiency.
Practice: the physical properties of digesta in relation to mixing
a) Viscosity
The apparent viscosity of herbivore digesta is high and non Newtonian in both the proximal and distal parts of the small intestine in spite of relatively low dry matter content of the former. The high apparent viscosity will inhibit the establishment of turbulence and promote laminar flow. This will occur even when there are sudden reversals of flow though the contracting segment as described by Jeffery et al. (2003), as high apparent viscosity will tend to decrease flow though the narrowed segment and to counter any sudden increase in flow velocity.
The non-Newtonian characteristics of herbivore small intestinal digesta will cause shear near the stationary wall of the intestine to promote a lowering of apparent viscosity. This effect will promote flow in the peripheral zone of the digesta occupying the lumen whilst the more viscid axial core will tend to move as a plug with uniform velocity (Silvester 1985). Similarly, shear thinning may augment flow though contracted segments of small bowel but will tend to damp down any turbulence in the lower shear conditions that prevail within the dilated segments lying proximal and distal to the contracted segment (Sweeney and Patrick, 1977; Walton et al., 1981).
b) Rheometry
The extent to which the solid phase may be elastically compressed and the degree to which whole digesta flows laterally on application of stress may be determined rheometrically as the ratio of elastic modulus (G') to the loss of viscous modulus (G'') i.e. G'/G''. This can be determined by placing the digesta in an annular ring on a rheometer and measuring the torque across the annulus when the central plug undergoes rapid alternate clockwise and anticlockwise rotation. This motion causes the sinusoid of strain to be 90○ out of phase with strain rate. Now the stress sinusoid of an elastic solid is in phase with deformation whereas the stress sinusoid of a viscous liquid is in phase with the strain rate. Thus by relating deformation to phase we can distinguish elastic modulus (G') from the loss or viscous modulus (G'').
The results of our rheometric analyses show that digesta from the proximal and distal small intestine of an herbivore behaves as a weak gel, the higher elastic modulus indicating an ability to sustain compression without flow. Thus, under appropriate physiological conditions, compression of digesta can occur causing the fluid phase to flow within the solid phase.
c) Permeametry
It remains to show that the hydraulic forces necessary for compression of the digesta plug lie within the physiological range of pressures that exist in the intestinal lumen. The range of hydraulic forces generated during intestinal propulsive and segmentative movements are variously reported between 2.5 cm (Gwynne et al., 2004) and 20 cm of water (Hightower, 1968). Our work, using static hydraulic pressures applied to fresh herbivore digesta in a vertically oriented permeameter, shows that significant compression of the gel structure and extrusion of the fluid phase from the solid phase of small intestinal digesta can occur at pressures lying within the reported physiological range.
d) Augmentation of plug flow
Given that extrusion of the fluid phase occurs in the area into which propulsive action is forcing digesta it can be hypothesised that a peripheral layer of less viscid fluid interposing between the viscid digesta plug and mucus layer could disengage any tendency of mucus to adhere to digesta (Denny, 1988) and thus expedite flow.
We tested this hypothesis by comparing the viscosity of herbivore small intestinal digesta determined with a capillary viscometer lined with a cylindrical section of fresh distal small intestine to that obtained using conventional rotatory viscometry. The viscosity profile of small intestinal digesta was again high and non-Newtonian, but the apparent viscosities were an order of magnitude below those obtained by rotatory viscometry, indicating that plug flow was being augmented during capillary flow. Such augmentation may be viewed as physiologically advantageous as it prevents trauma to the delicate intestinal mucosa during on-flow of digesta, but similarly could be viewed as disadvantageous in that by expediting flow it may decrease residence time. However the latter effect may be compensated to some extent by the physiological effects of luminal particulate matter which is reported to slow the rate of peristalsis (Larson and Schultz, 2002)
It is notable that ‘Hookean' elasticity of the solid phase of digesta may permit both egress and ingress of fluid from the digesta plug during phasic segmental peristalsis. Thus enzymatic components secreted by glandular elements within the intestinal walls (succus entericus) may be admixed with expressed fluid and drawn into the solid matrix of the digesta during lofting of the solid phase on relaxation following compression. Such admixed digestive enzymes would be re-expressed, along with soluble products of enzymatic digestion, in a subsequent contraction. In all, this process will augment the efficiency of digestion of the solid phase. However, this effect is likely to be reduced if intestinal compression exceeds the elastic limit of the matrix as this would lead to irreversible compaction with concomitant reduction of void spaces and permeability. Judging by the progressive increase in dry matter content of digesta from proximal to distal sites in the gut it seems likely that irreversible compression of digesta occurs during distal progression, presumably consequent on digestion reducing the strength of the matrix.
d) Inactivating the ‘no-slip' function of mucus
Some advances have shown that intestinal mucus has diverse functions (Allen et al., 1998) which include the formation of a microhabitat for the establishment of organisms that contribute to competitive exclusion of pathogens (Bourlioux, 1997) as well as forming a barrier to adherence of intestinal pathogens (Moncada et al., 2003). Thus optimal protection demands an intact mucus layer (Moncada et al., 2003) yet digestion demands close interaction of mucosa with digesta to achieve efficient digestion. The significant elastic properties of herbivore mucus would enable it to adhere to food particles (Denny, 1988) thus promoting mucosal adherence to the digesta plug which may delay its passage and promote digestion. One outcome of our results, showing augmentation of plug flow under hydrostatic pressure, is that the ‘no slip' effect of mucus may be disengaged during propulsive actions such a peristalsis and segmentation. As intestinal mucus forms a significant barrier to the diffusion of small uncharged molecules (Khanvilkar, 2001) it is likely that any fluid expressed from the periphery from the digesta plug will accumulate between the mucus layer and the digesta plug. This would provide a shear plane that effectively lubricates the plug during peristalsis regardless of the physical character of the epithelial surface.
Other factors modifying movement of the fluid phase of digesta
It is appropriate at this point to discuss other physical factors which may interact with the physical dynamics of digesta and influence digestive efficiency.
a) Particle size spectrum
Whilst feeds with finer particle sizes have a relatively greater surface area available for digestion they have smaller void spaces between adjacent particles when in packed arrays (Dullen, 1979). The latter effect results in a reduction of mean diameter of the pores available for permeation. In mixtures of large and small particles there may be relative movement of small particles within the void spaces between larger particles which may lead to progressive obturation of larger spaces during phasic fluid flow.
The effects of variation in particle size spectrum on digestibility are difficult to asses as the type of milling and setting up of the mill (Koch, 1996) along with the physical characteristics of the feed material that is being milled (Wright and Vincent, 1995) lead to differences in the shape, size, and distribution of particles that often preclude meaningful comparisons.
b) Viscosity of the fluid phase
The adverse effect of agents which increase fluid phase viscosity on digestibility is supported by extensive publications regarding the effectiveness of dietary gums in decreasing the efficiency of glucose and cholesterol absorption (Doi et al., 1979; Ebihara et al., 1981; Davidson et al., 1991).
If the apparent viscosity of the fluid phase contained within the digesta plug is increased by the presence of such agents, then the ease of expression of the fluid contained within the matrix during compression will be reduced (Dullen, 1979). This effect is likely in situations where steaming and subsequent pelleting of feeds has caused solublisation of starches and particles are agglomerated. Under such conditions any increase in digestibility from increase in surface area may be offset by local increases in the viscosity of the liquid phase.
References are available on request.
From Proceedings of the "17th Australian Poultry Science Symposium", New South Wales, Australia.
