| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ANIMAL NUTRITION |
,2
* Louisiana State University Agricultural Center, Baton Rouge 70803-4210 and
and
Texas A&M University, College Station 77843-2471
Abstract
A sequence of eight twice-daily meals, each marked with different rare earth elements, was fed to 24 Spanish goats (BW = 20.6 ± 1.94 kg) to produce meal-based profiles of rare earth markers within segments of the gastrointestinal digesta on subsequent slaughter. Accumulative mean residence time and time delay of rare earths and segmental and accumulative mean residence times of indigestible NDF (IDF) were estimated for each sampled segment. Diets consisted of ad libitum access to bermudagrass hay with a limit feeding of one of four supplements: 1) minerals (basal, B); 2) B + energy (E); 3) B + CP (CP); or 4) B + E + CP for 84 d. Mean daily intake (g/kg of BW) during the 5 d before slaughter differed (P < 0.05) via diet for DM but not for IDF (8.0 ± 0.35 g/kg of BW). Larger estimates of cumulative mean residence time for IDF vs. rare earths were suggested to be the consequence of a meal-induced bias in the single measurement of IDF pool size by anatomical site. The rare earth compartment method was considered more reliable than the IDF pool dilution method because it yielded flow estimates based on the flux of eight meal-dosed rare earth markers over 4 d and was independent of anatomical definitions of pool size. Statistically indistinguishable estimates for gastrointestinal mean residence times for IDF and rare earths conform to assumed indelibility for the specifically applied rare earths and indigestibility of IDF. The potentially digestible NDF (PDF):IDF ratio of dietary fragments (0.8) progressively decreased in the following order: caudodorsal reticulorumen (0.390) > crainodorsal reticulorumen (0.357) 
reticulum (0.354) > mid-dorsal reticulorumen (0.291) ventral reticulorumen (0.286), to that within the omasal folds and in the abomasum (0.259). Such a gradient of progressively aging mixture of plant tissue fragments is consistent with age-dependent flow paths established in the reticulorumen and flowing to the omasum and abomasum. Such heterogeneity of fragment ages within the reticulorumen is also indicated by the superior fit of marker dose site 
marker sampling site model assumptions. Additionally, cyclic meal- and rumination-induced variations in escape rate occur. Estimates of mean escape rates over days, needed for the practice of ruminant nutrition, must consider the complex interactions among plant tissues and the dynamics of their ruminal digestion of PDF.
Key Words: Digesta • Goats • Methodology • Rumen • Turnover
Introduction
The nutritive potential of meals is determined by the rate of escape (or conversely, the segmental mean retention time) and rates of digestion of nutritive entities of masticated fragments within segments of digesta. Various forces propel and constrain the flow of plant tissue fragments undergoing digestion within successive segments of the gastrointestinal tract of herbivores. Driven by nutrient deficits at the tissue level (Ellis et al., 1999
, 2000), continuing intake of sequential meals causes a flux of undigested residues being propelled by gastrointestinal motility through successive segments of the gastrointestinal tract. Various forces constrain this flux in different segments of digesta and cause differences in segmental loads and mean residence times of undigested fragments. Because of its diverticular arrangements of entry and exit flow (Moir, 1965) and well-organized and regulated motility cycles of the reticulorumen (Erhlein, 1980), primary emphasis has been directed to defining mechanisms that constrained flux of undigested fragments within the reticuloruminal digesta (Ellis et al., 1994). Quantitative estimates of flux through postruminal segments are lacking partly because of sampling problems in postgastric segments of digesta. This study used sequential dosing of multiple meal markers before slaughter to elicit a meal-based profile of markers within digesta segments sampled at a single time. Compartmental models were then fitted to these marker profiles to estimate mean residence time within each digesta segment. Postulated mechanisms constraining flux of each digesta segment were then evaluated.
Materials and Methods
This experiment was approved by the Animal Care and Use Committee of the Louisiana State University Agricultural Center. Twenty-four Spanish crossbred wether kids were fed Coastal bermudagrass hay-based diets for 84 d prior to this experiment. Before the 84-d period, the kids had been treated with an anthelminthic (Valbazen, Smith Kline and Beecham, Philadelphia, PA) and a coccidiostat (Amprolin, Aguet, Rahway, NJ). During both the 84-d and this experiment, the kids were penned individually in elevated pens (1.5 ± 1.5 m2) blocked within four physical locations in an open-sided building and allowed ad libitum access to Coastal bermudagrass hay.
Daily supplements to the hay (6.7% CP, 75.5% NDF, 36.4% ADF, and 6.1% ash, DM basis) were: 1) 47 g of basal minerals (B), 2) B + energy (E), 3) B + protein (CP), or 4) B + E + CP (ECP). The B supplement contained 10.4% urea, 23.0% dicalcium phosphate, 7.4% trace-mineralized salt, 4.3% molasses, and 54.9% ground corn on a DM basis. Also on a daily DM basis, E was supplied by 147.3 g of corn; CP by 56.3 g of corn, 68.4 g of corn gluten meal, and 27.6 g of fish meal; and EP by 193.3 g of corn, 48.3 g of corn gluten meal, and 27.6 g of fishmeal. The hay and supplement were fed at 0700 daily. Supplements were formulated to provide similar amounts of corn protein and fishmeal CP as in ECP, and similar energy as in E and ECP to prevent the possibility of CP and energy influencing performance. This resulted in different amounts of supplement fed.
Kids were denied feed and water for 16 h and weighed initially, at 28-d intervals, and at the end of the experiment. Kids averaged 20.6 ± 1.94 kg (mean ± SD) initially. One kid from each block was randomly assigned to each of the four treatments. Two kids assigned to receive supplement B would not eat the supplements and were removed from the experiment.
After completing the 84-d growth trial, digesta within the gastrointestinal tract of each kid was marked with a series of meals, each containing a different rare earth element. Kids were allowed 30 min to consume the partial meal of 50 g of marked hay. Hay unconsumed after 30 min was removed, weighed, and sampled for unconsumed marker and the additional allowance of unmarked hay was then fed. Average consumption of the marked hay was 46.4 ± 0.04 g. Meals of hay marked with Yb, Sc, Eu, Tb, Lu Sm, La, or Nd were offered at 72, 60, 48, 36, 24, 18, 10, and 6 h before slaughter. The order of rare earth meals was directly related to their sensitivity of analysis. However, the dose of the least sensitive Nd proved insufficient for reliable detection (twice background) in approximately 50% of the animals. Two or three kids from each treatment were assigned to receive the sequence of marked hay meals during two sequential 4-d periods. The first offering of marked hay (72 h preslaughter) within each period was staggered by 30 min for a target slaughter order of kids to ensure time to obtain individual animal data. Supplements were fed 1 h before the marked hay was fed at h 72, 48, and 24. Kids were dosed with 20 mL of 0.10 M cobalt diethylenetriaminopentaacetic acid (CoDTPA) at 48 h and 0.10 M chromium DPTA (CrDTPA) at 24 h. Marked hay and any unconsumed marked hay were collected, sampled, and stored for later analysis. Unmarked hay was offered after the marked orts were removed.
Marking was accomplished by soaking 2 kg of ground hay for 12 h in 2 L of deionized water containing 1 g of each individual rare earth as the chloride salt. The hay was rinsed with tap water, with 0.01 M acetic acid, and then again with tap water to remove unbound rare earth prior to drying at 60°C. Daily intake of each diet was measured over a 5-d period before slaughter.
Beginning 6 h after the morning feeding, each kid was killed at 30-min intervals (in the targeted slaughter order of meal feeding) by captive bolt immobilization and jugular exsanguination. The abdominal cavity was opened, and the gastrointestinal tract was removed and ligated to prevent further mixing of digesta among sampling sites. Digesta samples were collected and weighed from the ventral rumen and cranial, mid, and caudal sites of the dorsal rumen. The remaining reticuloruminal digesta was removed, weighed, thoroughly mixed, and sampled.
All digesta was removed and weighed from the reticulum, omasal folds, mid-omasum, abomosum, small intestine, cecum, colon, and rectum. Digesta was frozen and later freeze-dried before grinding to pass a 2-mm screen for subsequent analysis. Samples of feed, refused feed, and freeze-dried digesta were analyzed for DM, ash, CP (AOAC, 1990), and NDF (Goering and Van Soest, 1970). Indigestible NDF (IDF) was determined as NDF remaining after a 144-h in situ incubation (Lippke et al., 1986) in bags of 20- to 30-µm porosity in the rumen of a cow fed the bermudagrass hay. Samples of marked hay, unconsumed marked hay, and dried digesta were analyzed for rare earth elements by neutron activation analysis (Pond et. al., 1985) and for IDF. Potentially digestible NDF (PDF) was computed as the difference between NDF and IDF in each sample.
Digesta Flow Models
External Markers Dosed as Meals. Dose of each rare earth element by individual animal was computed as the quantity of marker in hay offered minus the marker in hay refusal. Concentrations of each rare earth element in digesta from each animal were adjusted for differences in quantity of rare earth actually consumed based on measured concentrations of each element bound and the quantity of meal consumed. Such equivalent concentrations expected for each rare earth element meal at dose time t (g) was computed by individual animal as actual consumption of element with meal t, g, x (actual dose of element with meal t, g,/actual dose of element with meal 72 h before slaughter). Compartmental models were then fitted to profiles of equivalent marker concentrations over the lapse time of rare earth meals using PROC NLIN of SAS (SAS Inst., Inc., Cary, NC).
The compartmental flow models fitted to profiles of the eight rare earth elements have been described by Ellis et al. (1994). The G3
G1
O model was found to be most appropriate for fitting the data (Figure 1). In the G3G1O model, the initial mixing pool encountered by the marker is characterized by a G3 age-dependent turnover (G3) to the second mixing pool characterized by age-independent turnover (G1) to a pool characterized by nonmixing laminar flow or time delay () before the marker appears in the efflux from the digesta segment. The mean residence time due to compartmental mixing (CMRT) in the initial entry, G3 pool (CMRT1), was computed as 3/
1. The CMRT in the second mixing pool (CMRT2) was computed as 1/k2. The accumulative mean residence time (AMRT) for CMRT turnover and laminar flow () in the digesta segment was then computed as the sum of CMRT1 + CMRT2 + .
|
. The approximate AMRT for solutes was estimated as AMRT = (lnCr in digesta segment - lnCo in digesta segment)/(dose time for Cr - dose time for Co). Such approximation assumes that any concurrent G1 was negligible within the time interval of 48 to 24 h of measurement.
Mean Residence Times.
The segmental mean residence time (SMRT) of the dietary (hay, corn and soybean meal) IDF was computed as SMRT = (g of IDF in the segment of digesta)/(g/d flux rate of IDF to the feces during the 5 d before slaughter). The SMRT of IDF represents the mean residence time of IDF pool in the segment of digesta measured without assuming any flow mechanism. In contrast, the external marker was administered as discrete doses that were diluted with digesta in successive flow segments preceding the digesta sampling site. Thus, fitting compartmental models to the flux of a single dose of external markers within successive segments of digesta estimates the AMRT within all preceding segments of digesta. The AMRT estimated by the compartmental model represents the accumulative mean residence time due to the sum of mean residence times specifically due to mixing (CMRT1 and CMRT2) and nonmixing laminar flow (), or for the entire gastrointestinal tract, CMRT1 + MRT2 + .
Dose Site vs. Sample Site Models.
Flow through an age-dependent mixing pool or through two sequential mixing pools cannot be modeled without knowledge of dose and sampling sites. Models G3 and G1 in Figure 1
assume marker dose site = marker sample site and estimate marker remaining in the site. Because digesta turnover from the sum of G3 and G1 pools are commingled and cannot be separately sampled as conceptualized in Figure 1, turnover from the composite of the two pools was estimated by the G3G1 model (Figure 1). Alternatively, sampling the efflux (O) from the two pools reflects net turnover of digesta emerging from the commingled system of flow mechanisms and can be estimated by model formulation G3G1O (Figure 1).
The model G3G1O (Figure 1) partitions laminar residence time, , from nonlaminar residence time. Laminar flow is defined as polar flow from entry to exit without turbulences or mixing and results in a discrete time delay (Ellis et al., 1994).
Statistics
Data for apparent digestibility of dietary components and intake of dietary components were analyzed as a randomized complete block designed experiment with treatments arranged as a 2 x 2 factorial. Terms of the model included block, level of energy, level of CP, and the interaction of level of energy and level of CP. Missing values for two animals (in the same block) were treated as programmed for missing values in PROC GLM of SAS. Data for load of IDF, SMRT, and AMRT were analyzed by sampling site as a randomized complete block design with treatments arranged as a 2 x 2 factorial. Terms in this model included block, level of energy, level of CP, and the interaction of level of energy and level of CP. Data for the mean residence times associated with compartmental mixing, time delay, and AMRT of both rare earth markers and IDF were analyzed as a completely randomized designed experiment using a model that included terms for the site of digesta sampling within the gastrointestinal tract. For these data, means were separated using Fisher’s protected LSD and a probability of P < 0.05 when the F-test was significant. The AMRT for the total gastrointestinal tract was calculated using both rare earth markers and IDF. These data were then combined into a single dataset and analyzed as a completely randomized designed experiment. The model included only a term for the method (rare earth or indigestible fiber) used to make the estimate. Because there are missing data, all data are presented as adjusted least squares means, and the largest standard errors associated with any one variable are presented. All calculations were completed using SAS.
The error mean square (EMS) was used as the criteria of model fit to data. The Wilcoxon’s signed rank procedure (Ott, 1977) was used to test the hypothesis of similar distributions of EMS as estimated by each model for various fractions. If two models were found to have indistinguishable EMS distributions, they were regarded as being equal in their fit to the data. In cases where the null hypothesis of similar EMS distribution of two models was rejected, the model with a smaller median EMS was considered superior.
Results
Intake and Utilization of Diets
Ad libitum hay diets selected by the kids ranged from 49 to 88% hay (Table 1). Apparent digestibility of PDF decreased, whereas that of CP increased as CP in the diet increased (P < 0.05, Table 1). Apparent digestibility of CP decreased (P <0.05) when level of E was increased, but digestibility of PDF was not affected. There were no interactions between level of E and CP for digestibility of PDF or CP (Table 1). Intake rates of NDF and IDF were associated with interactions (P < 0.01) between CP and E (Table 2).
|
|
. The dominant effects of level of CP and E were associated within digesta of the omasal folds, small intestines, colon, and rectum, where levels of both CP and E decreased (P < 0.06) load of IDF without significant interactions. In contrast, load of IDF in digestive segments, presumably involved in most active fermentation (reticulorumen and cecum), were unrelated to CP and E, with the exception of effects of E (P = 0.02) and its interaction with CP (P < 0.03) in the cecum. Level of CP, but not energy, had significant (P > 0.05) effects on gastrointestinal load of IDF and in increasing reticulorumen load as a proportion of that in the total gastrointestinal tract.
|
. Quantitatively, the dominant dietary effect was that of level of E on increasing (P < 0.01) SMRT in the reticulorumen, as shown in Table 4, or when expressed as a function of BW or as a proportion of accumulative mean residence of IDF in the total gastrointestinal tract (data not shown). The SMRT of IDF appears to be initially established in the reticulorumen digesta, which also contains the dominant proportion of the total gastrointestinal tracts’ mass (0.62 to 0.8, Table 3) and SMRT (0.65 to 0.77, Table 4) of IDF. As a consequence of this dominance within the reticulorumen, effects of E on SMRT in initial flow segments of the reticulorumen appear causal of effects of E on SMRT in subsequent, postreticuloruminal flow segments.
|
through G6 to the data minimized RMSE for G = 4. Therefore, the two-compartment model equivalent to the one-compartment, age-dependent model of G4 (the G3G1 model; Pond et al., 1988) was chosen for use in estimating flow parameters.
In the present data, both the G3G1 (dose site path = sampling site path) and the G3G1O (dose site model sampling site model) were fitted to segments of the ruminal digesta to evaluate model assumptions of marker dose and marker sampling site. The model assuming marker dose site marker sampling site provided superior statistical fit (P < 0.01) for all subsegments of ruminal ingesta sampled, as well as for postruminal sites. The superior fit of the marker dose site marker sampling site vs. the model of marker dose site = marker sample site with (G3G1O) is illustrated for one kid in Figure 2. It was concluded that parameters estimated via the G3G1O model provided statistically superior estimation and parameters estimated by this model will be used in subsequent discussions.
|
. Dietary effects (P < 0.01) only of E on AMRT occurred in the rectum, with interactions between CP and E (P < 0.08) occurring at all sites other than the small intestines and reticulorumen.
|
(Figure 1) and SMRT for IDF and approximated AMRT of solutes are compared by sampled digesta site in Table 6. Clearly, AMRT for solutes was extremely unresponsive to dietary treatment and effects of digesta site effects, as was IDF and rare earths. Consistent with physiological expectations, AMRT of rare earths and SMRT of IDF increased progressively through more caudal segments flow segments of digesta as ordered in Table 6. With minor exception, the AMRT of rare earths were essentially established in the reticuloruminal digesta, increased slightly in postruminal and prececal segments (1.25-fold), and increased precipitously in cecal and postcecal segments of the digesta (1.8-fold). Increases in both CMRT1 and CMRT2 with more caudal progression of digesta accounted for the major portion of the progression of AMRT. The exhibited very slight progression (Table 6).
|
). In contrast, the AMRT for IDF in the reticuloruminal digesta accounted for 72% of the AMRT for the total tract (37.4 h AMRT in the reticuloruminal digesta/52.1 h AMRT through the entire tract and through the rectum, Table 6). The AMRT in the total gastrointestinal tract did not differ (P = 0.45) for the rare earths vs. IDF (56.3 vs. 53.5 h, Table 6). The AMRT for solute markers (CoDTPA and CrDTPA) exhibited only a small progression with segment of digesta being essentially unchanged in precaecal segments and increasing slightly (approximately 1.2-fold) in postcaecal segments of digesta.
Discussion
Unique results of this study include 1) equivalent AMRT for IDF and rare earth elements through the rectum (52.1 vs. 57.3 h, Table 6), but a slower AMRT for IDF than for rare earths through reticuloruminal digesta (37.4 vs. 23.1 h, Table 6) and 2) marker profiles of rare earths in ruminal digesta inconsistent with dose site = sampling site (Figures 1 and 2). We propose that such disparate results for rare earths vs. IDF are due to timing of a single sample to estimate reticuloruminal pool size of IDF in combination with the dynamic and heterogeneous mixture of age-dependent flow paths of undigested residues through reticuloruminal digesta.
Gastrointestinal AMRT for Rare Earths vs. IDF.
Previous studies have reported equivalent mean gastrointestinal AMRT for IDF and for specifically applied rare earths in sheep (Wylie et al., 1986; Ellis et al., 2002) and cattle (Ellis et al., 1994; Huhtanen and Vanhatalo, 1997). Faichney et al. (1989) reported that specifically applied rare earth elements yielded AMRT in ruminal digesta similar to IDF. The calculation of dilution flow parameters for the two types of markers (IDF and specifically applied rare earths) involves identical assumptions, the most important of which include 1) steady-state conditions with respect to feed intake and digesta load and flow, 2) indelibility of external markers (Worley et al., 2002), and 3) no effect of bound rare earths on properties of plant tissue fragments affecting their flow (such as specific gravity).
We propose that the major difference between estimates of flow parameters by the two methods resides in estimates of the resident mass in the reticulorumen that dilutes the flux of IDF at one point in time (2 h after AM meal) vs. the diluting flux of 8 different rare earths over 4 d. As used here, the IDF dilution method estimates the mean dilution rate of IDF, IDFke, as mean flux of IDF during a 4-d period divided by resident mass of IDF at the end of the 4-d period in the anatomically defined digesta mixing pool 2 h after the AM meal. In contrast, as used here, the rare earth compartmental model method estimates the mean diluting rates of the rare earths, REke, from changes in concentration of rare earth elements fluxing through the sampling site. Estimates of REke are derived from a series of eight different rare earth elements dosed as serial meals over 4 d, whereas IDFke is derived from a single estimate of resident mass occurring 2 h after completion of the last meal. Also, the IDF dilution method assumes an anatomically definable mixing pool, whereas the compartmental model method pool size(s) are established by the assumed marker-digesta mixing mechanism(s) (e.g., pool size = the mass in which the introduced marker achieves equilibration).
In retrospect, we propose that estimates of IDF pool size 2 h after feeding was larger than the mean daily IDF pool size associated with the mean daily flux of IDF and this resulted in an overestimation of AMRT of IDF. In contrast, the serial flux of eight rare earth elements in eight serial meals results in estimates of a mean REke representing concurrent flux of marker and diluting mass from eight meals over 4 d. With representative measurements of IDF pool size, the IDF pool dilution method yields results consistent with the rare earth marker compartmental model method (Ellis et. al., 2000).
Consequently, we propose that the scale of parameters estimated by the IDF dilution model as implemented here is biased and must be interpreted only in relative terms. Further implications of nonsteady-state process in estimating flow will be emphasized and discussed later.
AMRT for IDF vs. Undigested PDF.
It is frequently speculated that escape of undigested PDF and IDF entities may differ and such inequalities will limit conclusions assuming equality of escape of undigested PDF and IDF. Such speculation overlooks the fact that undigested PDF and IDF are both insoluble components of fragments of the same tissues and fragments thereof and, until digested, must have escape rates identical to the mixture of undigested fragments of plant tissues they comprise. The entities of PDF and IDF are not determined in discrete fragments but rather in a distribution of fragments having a distribution of physical and chemical attributes as well as flow rates. It is the mean of this distribution of fragments that is estimated from the dilution of the mass (or pool) of an indigestible entity such as IDF (mean dilution rate of IDF/h = flux of IDF, g•h-1•g of IDF in pool-1) or an indelible marker attached to the flux of the fragment (flux of entity, g/h/pool of marker, g). Fragments from different plant tissues may have different proportions of PDF and IDF, and differential distributions of their flow will result in different distributions of PDF/IDF proportions. Only by assuming undifferentiated distributions of fragments in their ratio of PDF:IDF and steady-state conditions and knowing the rate of digestion of PDF allows estimation of escape rate of undigested PDF (Ellis et al., 2000; Lund et. al., 2003).
Changes in PDF:IDF of a mixture of fragments undergoing digestion can be used to infer effects of age on digestion of PDF (decreasing PDF:IDF) and/or on a redistribution of fragments with respect to different PDF:IDF ratios (increasing or decreasing PDF:IDF). In order to evaluate such distributions in the reticuloruminal digesta, PDF:IDF ratios were computed and are summarized in Table 7.
|
reticulum (0.354) > mid-dorsal reticulorumen (0.291) ventral reticulorumen (0.286), to that within the omasal folds and in the abomasum (0.259). Such a gradient of progressively aging mixture of plant tissue fragments is consistent with flow paths observed for aging fragments in the reticuloruminal digesta (Ehrline et al. 1980). However, considering the dynamics of fragments in reticuloruminal digesta (see Figure 3 and discussion later), it is surprising that such discrete intraruminal site differences could be maintained and statistically detected.
|
Profiles of PDF:IDF in gastric segments (Table 7) suggest digestion-dependent (age-dependent) flow paths of rare earths are established in the reticuloruminal digesta and, with the exception of the small intestine, are extended to segments of the large intestines before precipitous increases in PDF digestion in the colon and cecum. Such gastrointestinal profiles of PDF:IDF coincide with profiles of AMRT for the rare earths that also precipitously increased in the colon and cecum (Table 6). The extent of digestion of PDF observed here in goats contrasts sharply with the relatively small postgastric AMRT observed in cattle (Ellis et. al., 1994; Huhtanen and Kukkonen, 1995; Wylie et al., 2000).
Ruminal Dynamics of Feed Fragments.
The apparent systematic differences in PDF:IDF of fragments within anatomically defined segments of the reticuloruminal digesta (Table 7) clearly indicates that mixing of fragments entering and flowing through reticuloruminal digesta is neither perfect nor achieves a random distribution of uniformly aged fragments (i.e., digestion of PDF:IDF). Slow and imperfect mixing of younger fragments on initial entry into the reticuloruminal digesta can be accommodated by the age-dependent flow rate modeled as a gamma 3 (G3) turnover process (Figure 1). Escape from the initial G3 mixing compartment (e.g., pool) to the second compartment is modeled as mass action kinetics (G1 turnover process). Although modeled as discrete and sequential mixing compartments, each conceptual compartment cannot be physically identified and specifically sampled. Due to the unceasing mixing of reticuloruminal digesta by the cyclic ruminal motility, the two pools are physically commingled. The two compartments are expressed and estimated by their physical representation of intrareticuloruminal flow processs.
Assuming rapid mixing of the two sequential pools, the digesta dose site for the external marker (rare earths) would be indistinguishable from the digesta sampling site used to determine marker dilution (turnover) rate. With perfect mixing of marker introduced into reticuloruminal digesta, the initial dilution of marker at time of introduction would yield some maximal concentration at dose time followed by a cumulative turnover from the two pools (G3G1 or ruminal dose site = ruminal sampling site model, Figure 1). In contrast, with slow and imperfect mixing, such that dose site sampling site, marker flow through the two sequential initial compartments would be delayed and emerge from the reticulorumen (outflow site = O) as ascending concentrations from zero at marker dose time (G3G1O or ruminal dose site rumen outflow sampling site).
As inferred by the systematic variations in PDF:IDF of fragments in reticuloruminal digesta (Table 7) considerable variations occurs in sampling age-dependent processes of ruminal flow. Consequently, estimates of flow from reticuloruminal digesta are estimated with greater sensitivity by fitting model G3G1O to marker dilution profiles emerging the rumen.
Marker Dose and Sampling Site Models.
Although marker profiles in reticuloruminal digesta generally conformed to expectations of marker dose site = marker sampling site (Figure 2), statistical analysis indicated a superior fit (P < 0.001) for the marker dose site marker sampling site for all sampling sites of digesta including reticuloruminal digesta. An insufficient dose of Nd 6 h before slaughter resulted in low and undetectable concentrations of this marker in many animals and consequent uncertainty in partitioning residence time for the G3 pool and its resolution from . A few (9 out of a total of 144) samples exhibited an ascending value at 6 h (see VR and mixed ruminal digesta for kid 35, Figure 2).
In conclusion, the statistical superiority of the marker dose site sampling site model clearly indicates the presence of two mixing compartments commingled in reticuloruminal digesta. Because of slow mixing and diural- (Deswysen et al., 1989b) and meal-based variations in SMRT within digesta of the reticulorumen is most appropriately estimated by fitting the marker dose site marker sampling site model to digesta emerging the reticulorumen.
Partition of Gastrointestinal SMRT.
Previous studies involving intestinally cannulated ruminants have indicated the reticuloruminal digesta as the major site of SMRT of feed fragments in cattle (Ellis et al., 1994; Huhtanen and Vanhatalo. 1997; Wylie et al., 2000) and sheep (Ellis et al., 2003). The dominant SMRT for undigested residues in the reticulorumen (0.65 to 0.77 of gastrointestinal SMRT, Table 4) suggests that the relatively small contributions of the hindgut may be characteristic of Spanish goats in contrast with other small ruminants. Hoffman (1988) notes the increased load of digesta in the hindgut compared to the reticuloruminal digesta of selective grazers (such as goats) compared to that of indiscriminate grazers (such as cattle) and suggested that increased load of digesta in the hindgut is an evolutionary characteristic of selective grazers in general. Faichney (1984) reported comparable SMRT for the hindgut compared with the reticulorumen of pregnant sheep. This latter observation in sheep may be related to space competition of the products of conception with space occupied by digesta in the reticulorumen and may be a factor that contributes to the SMRT of the hindgut. Alternatively, as suggested by Hoffman (1988), the hindgut of selective grazers such as goats and sheep may involve more extensive digestion of PDF (Table 7) and longer AMRT than that of indiscriminate grazers such as cattle.
The proportions of reticuloruminal CMRT1 to CMRT2 observed here for goats (1.3/19.4 h or 0.068 for mixed reticuloruminal digesta, Table 6) appear small relative to that reported for foraging cattle (0.29, Wylie et al., 2000; 0.34, Ellis et al., 1994; 0.15, Huhtanen and Kukkonen, 1995). In foraging cattle and sheep, CMRT1 and appear rather constant in the order of 8 to 12 h, and CMRT2 increases from 10 to nearly 40 h with increasing dietary concentrations of IDF (Ellis et al., 1999). It should be noted that the goats in this study consumed mixed diets ranging from 49 to 88% hay (Table 1). Previously, gastrointestinal SMRT has been altered by level of concentrate (Wylie et. al., 1990). However, diets of comparable proportions of hay and concentrates fed to cattle had much greater proportions of CMRT1/CMRT2 and larger (Huhtanen and Kukkonen, 1995).
Another difference in the results of this study is a very short of 1 to1.9 h compared with typical values ranging from 8 to 10 h for foraging cattle (Wylie et al., 2000) and sheep (Ellis et al., 2003). Wylie et al. (2000) has emphasized the difficulty of identifying the discrete digesta sites of the nonlaminar flow estimated by compartmental models. As estimated by compartmental models, may or may not be physiologically equivalent to conceptual laminar flow (Ellis et al., 1994).
A mixing compartment is defined by a volume in which input (marker) is instantaneously mixed with all resident fragments within that volume. Ingestively masticated fragments are slowly and incompletely mixed with resident residues of prior meal fragments undergoing digestion in the reticulorumen (Ehrlein, 1980). Incorporation of age dependency in the initial mixing pool mimics slow mixing/equilibration (France et al., 1990) and may be considered a distribution of age-dependent digestion lifetimes (Matis, 1972) or, conversely, a distribution of discrete digestion lags (France et al., 1985; Ellis et al., 1994). Increasing order of age dependency results in increasing SMRT partitioned to CMRT1 at the expense of (Matis, 1972; Pond et al., 1988).
In conclusion, goats, like ruminants in general, have varied flow processes impeding the flow of feed fragments through gastrointestinal segments. In these goats, there is not a clear distinction of sites contributing distributed lags (age-dependent mixing, CMRT1) from discrete lags (). The most effective retardation of flow appears to that of mass action competition among fragments for escape (CMRT2) from the reticulorumen. Processes impeding escape of fragments appear to be established in critical segments of reticuloruminal digesta and maintained caudal to the colon/cecum where further impedance occurred. Independence of escape from the reticulorumen involves forces deriving from both physical and chemical processes (digestion) in that these processes and their ensuing forces are anatomically commingled. Like hindgut-fermenting herbivores (Cork, et al., 1999), selective partitioning of fragment more abundant in PDF from the colon to the cecum appears causal of the precipitous declines in PDF/IDF in the colon and cecum of these goats.
Intermeal Variations in AMRT.
Ruminants consume feed as several main meals and numerous minor meals per day. Ruminal escape of smaller, more extensively ruminatively masticated fragments (Reid et. al., 1979) and DM is accentuated during major meals (Aitchison et al., 1986a,b; Gill et al., 1999). Lack of large fluctuations in reticuloruminal load during intermeal intervals (Gasa et al., 1991; Thiago et al., 1992; Huhtanen and Vanhatalo, 1997) suggests that inflow of DM generally exceeds the accentuated DM outflow during eating.
Figure 3 is an attempt to portray intermeal variations in escape of masticated fragments from reticuloruminal digesta as three recurring phases with respect to the abundance of differently sized fragments produced by ingestive and ruminative mastication. Such meal-based phases of fragment escape are combined with earlier concepts that describe the balance of physical forces partitioning flux of fragments into "lag rumination" and "mass action turnover pools" (Ellis et al., 2000). Additionally, the concept of "flow paths" is substituted for mixing pools (Jacqez 1996; Matis and Kiffe, 2002), as illustrated in Figure 1.
Intrinsic buoyancy of larger-sized ingestively masticated fragments and remnants of their constituent tissues positions these fragments into the lag-rumination flow paths. Colonization and fermentation of microbial accessible PDF results in a fermentation-based buoyancy that sustains larger fragments in lag-rumination flow paths. With a reduction in size, fragments are less effectively comminuted, and microbial accessible PDF is depleted. Partitioning of fragments is achieved as a balance between the mass action dilution-mixing forces of the primary mixing cycles of motility of the reticulorumen motility opposing the unmixing forces of buoyancy of individual fragments is achieved. Balance between these mixing forces sustains flow of "younger" fragments in the lag-rumination flow paths. With age in the lag-rumination flow path, fragments loose buoyancy sufficiently to escape to the "mass action" turnover pool where their escape is related to the diluting mass of fragments largely depleted of microbial accessible PDF and less buoyant. Thus, age, size, and buoyancy of fragments are all interrelated, but buoyancy and mass action competition among fragments are the opposing forces establishing flow paths through the reticuloruminal digesta and, finally, the selective ruminal escape of smaller, PDF-depleted, less buoyant fragments. The physical characteristics of fermentation-based buoyancy position individual fragments within a buoyancy-based gradient of fragments (unmixing forces). It is the mixing forces of the incessant and cyclic mixing forces of the coordinated reticuloruminal motility that perturbs the fermentation based gradients and imparts diffuse boundaries to motility sequences leading to "lag rumination" and motility sequences leading to "mass action turnover" flow-paths, respectively. Presumably, it is the accentuated, meal-induced primary sequences of motility that specifically accentuate ruminal escape from the "mass action" pool.
Finally, intraday variations in plant tissue fragment escape appear initiated by diurnal variations in apportioning of time spent eating vs. rumination (Deswysen et al., 1989a, 1993).
Implications
Mechanisms constraining ruminal flux of plant tissue residues seem highly probabilistic, and current methods based on markers and flow models must be considered as mean estimates over the time period involved.
Footnotes
1 Approved by the Director of the Louisiana Agric. Exp. Stn. as publication No. 02-11-0115. 
2 Correspondence: 2471 TAMU (phone: 979-845-5063; fax: 979-845-5292; e-mail: w-ellis{at}tamu.edu).
Received for publication October 3, 2002. Accepted for publication October 14, 2003.
Literature Cited
Aitchison, E. M., M. Gill, M. S. Dhanoa, and D. F. Osbourn. 1986a. The effect of digestibility and forage species on the removal of digesta from the rumen and the voluntary intake of hay by sheep. Br. J. Nutr. 56:463–476.[Medline]
Aitchison, E. M., M. Gill, and D. F. Osbourn. 1986b. The effect of supplementation with maize starch and level of intake of perennial ryegrass (Lotium perenne cv. Endura) hay on the removal of digesta from the rumen of sheep. Br. J. Nutr. 56:477–486.[Medline]
AOAC. 1990. Official Methods of Analysis. 15th ed. Assoc. Offic. Anal. Chem., Arlington, VA.
Cork, S. J., I. D. Hume, and G. J. Faichney. 1999. Digestive strategies of nonruminant herbivores: The role of the hindgut. Pages 210–260 in Proc. 5th Int. Symp. on the Nutr. of Herbivores. H. G. Jung and G. C. Fahey, Jr., ed. Am. Soc. Anim. Sci., Savoy, IL.
Deswysen, A. G., P. A. Dutilleul, and W. C. Ellis. 1989a. Quantitative analysis of nycterophemeral eating and ruminating patterns in heifers with different voluntary intakes and effects of monensin. J. Anim. Sci. 67:2751–2761.
Deswysen, A. G., P. Dutilleul, J. P. Godfrin, and W. C. Ellis. 1993. Nycterophemeral eating and ruminating patterns in heifers fed grass or corn silage: Analysis by finite Fourier transform. J. Anim. Sci. 71:2739–2747.[Abstract]
Deswysen, A. G., K. R. Pond, E. Rivera-Villarreal, and W. C. Ellis. 1989b. Effects of time of day and monensin on the size distribution of particles in digestive tract sites of heifers fed corn silage. J. Anim. Sci. 67:1773–1783.
Ellis, W. C., J. H. Matis, T. M. Hill, and M. R. Murphy. 1994. Methodology for estimating digestion and passage kinetics of forages. Pages 682–756 in Forage Quality, Evaluation, and Utilization. G. C. Fahey, ed. University of Nebraska, Lincoln.
Ellis, W. C., D. P. Poppi, J. H. Matis, H. Lippke, T. M. Hill, and F.M. Rouquette, Jr. 1999. Dietary-digestive-metabolic interactions determining the nutritive potential of ruminant diets. Pages 423–481 in Proc. 5th Int. Symp. on the Nutr. of Herbivores. H. G. Jung and G. C. Fahey, Jr., ed. Am. Soc. Anim. Sci., Savoy, IL.
Ellis, W. C., D. Poppi, and J. H. Matis. 2000. Feed intake in ruminants: Kinetic aspects. Pages 335–363 in Farm Animal Metabolism and Nutrition: Critical Reviews. J. P. F. D’Mello, ed. Commonwealth Agricultural Bureaux Int., Oxon, U.K.
Ellis, W. C., M. J. Wylie, and J. H. Matis. 2002. Validity of specifically applied rare earth elements and compartmental models for estimating parameters for flow of undigested plant tissue residues through the gastrointestinal tract of ruminants. J. Anim. Sci. 80:2753–2758.
Erhlein, H. J. 1980. Forestomach motility in ruminants. Film C 1328 des IWF, Goettingen, W. German. Pages 1–29 in Publikationen zu Wissenschagtlichen Filmen, Sektion Medizin, Serie 5, No. 9/C 1328.
Faichney, G. J. 1984. The kinetics of particulate matter in the Reticulorumen. Pages 173–196 in Control of Digestion and Metabolism in Ruminants. L. P. Milligan, W. L. Grovum, and A. Dobson, ed. Prentice Hall, Englewood Cliffs, NJ.
Faichney, G. J., C. Poncet, and R. C. Boston. 1989. Passage of internal and external markers of particulate matter though the reticulo-rumen of sheep. Reprod. Nutr. Dev. 29:325–339.
France, J., J. H. Thornley, M. S. Dhanoa, and R. C. Siddons. 1985. On the mathematics of digesta flow kinetics. J. Theor. Biol. 113:743–758.[Medline]
France, J., J. H. M. Thornley, R. C. Siddons, and M. S. Dhanoa. 1990. Determination of Reticulo-rumen volume in non-steady state using digesta markers: A dynamic, deterministic, mechanistic approach. Pages 159–168 in Modeling Digestion and Metabolism in Farm Animals. A. B. Robson and D. P. Poppi, ed. Lincoln Univ. Press, Canterbury, New Zealand.
Gasa, J., K. Holtenius, J. D. Sutton, M. S. Dhanoa, and D. J. Napper. 1991. Rumen fill and digesta kinetics in lactating Friesion cows given two levels of concentrate with two types of grass silage ad lib. Br. J. Nutr. 66:381–398.[Medline]
Goering, H. K., and P. J. Van Soest. 1970. Forage Fiber Analysis. Apparatus, Reagents, Procedures and Some Applications. Agric. Handbook No. 379, ARS-USDA. Washington, DC.
Gill, M., P. H. Robinson, and J. J. Kennelly. 1999. Diurnal patterns in rumen volume and composition of digesta flowing to the duodenum. Anim. Sci. 69:237–249.
Hoffman, R. R. 1988. Anatomy of the gastrointestinal tract. Pages 14–43 in the Ruminant Animal, Digestive Physiology and Nutrition. Prentice Hall, Englewood Cliffs, NJ.
Huhtanen, P., and U. Kukkonen. 1995. Comparison of methods, markers, sampling sites and models for estimating digesta passage kinetics in cattle fed at two levels of intake. Anim. Feed Sci. Technol. 52:141–158.
Huhtanen, P., and A. Vanhatalo. 1997. Ruminal and total plant cell-wall digestibility estimated by a combined in situ method utilizing mathematical models. Br. J. Nutr. 78:583–598.[Medline]
Jacquez, J. A. 1996. Compartmental Analysis in Biology and Medicine. 3rd ed. Bio Medware, Ann Arbor, MI.
Lippke, H., W. C. Ellis, and B. F. Jacobs. 1986. Recovery of indigestible fiber from feces of sheep and cattle on forage diets. J. Dairy Sci. 69:403–412.
Lund, P., M. R. Weisbjerg and T. Hevelplund. 2003. Estimation of mean ruminal retention time of DNDF in dairy cows based on combined data from rumen evacuation and marker excretion curves. J. Anim. Sci. 81(Suppl. 1):341–342 (Abstr.).
Matis, J. H. 1972. Gamma time-dependency in Blaxter’s compartmental model. Biometrics 28:597–602.
Matis, J. H., and T. R. Kiffe. 2002. On the use of time lags and time-dependent transfers in compartmental systems. Pages 109–137 in Advances in Stochastic Modelling. J. R. Artalejo and A. Krishnamoorthy, ed. Notable Publications, Nesharic Station, NJ.
Moir, R. J. 1965. The comparative physiology of ruminant-like animals. Pages 1–23 in Physiology of Digestion in the Ruminant. R. W. Dougherty, ed. Butterworths, Washington, DC.
Ott, L. 1977. An introduction to statistical method and data analysis. Duxburry Press, North Situate, MA.
Pond, K. R., W. C. Ellis, W. D. James, and A. G. Deswysen. 1985. Analysis of multiple markers in nutrition research. J. Dairy Sci. 68:745–750.
Pond, K. R., W. C. Ellis, J. H. Matis, H. M. Ferreiro, and J. D. Sutton. 1988. Compartment models for estimating attributes of digesta flow in cattle. Br. J. Nutr. 60:571–595.[Medline]
Pond, K. R., W. C. Ellis, J. H. Matis, and A. G. Deswysen. 1989. Passage of chromium-mordanted and rare earth labeled fiber: Time of dosing kinetics. J. Anim. Sci. 67:1020–1028.
Poppi, D. P., W. C. Ellis, J. H. Matis, and C. E. Lascano. 2001. Marker concentration patterns of labelled leaf and stem particles in the rumen of cattle grazing bermudagrass (Cynodon dactylon) analysed by reference to a raft model. Br. J. Nutr. 85:553–563.[Medline]
Reid, C. S. W., A. John, M. J. Ulyatt, G. C. Waghorn, and L. P. Milligan. 1979. Chewing and the physical breakdown of feed in sheep. Ann. Rech. Vet. 10:205–207.[Medline]
Sutherland, T. M. 1989. Particle separation in the forestomach of sheep. Pages 43–73 in Aspects of Digestive Physiology in Ruminants. A. Dobson, and M. H. Dobson, ed. Cornell Univ. Press, Ithaca, NY.
Thiago, L. R. L., M. Gill, and J. W. Sissons. 1992. Studies of method of conserving grass herbage and frequency of feeding in cattle. Br. J. Nutr. 67:319–336.[Medline]
Worley, R., A. Clearfield, and W. C. Ellis. 2002. Binding affinity and capacities for ytterbium (3+) and hafinum (4+) by chemical entities of plant tissue fragments. J. Anim. Sci. 80:3307–3314.
Wylie, M. J., W. C. Ellis, and J. H. Matis. 1986. Validity of rare earths as flow markers for undigested feed residues. J. Anim. Sci. 63 (Suppl. 1):5. (Abstr.)
Wylie, M. J., W. C. Ellis, J. H. Matis, E. M. Bailey, W. D. James, and D. E. Beever. 2000. The flow of forage particles and solutes through segments of the digestive tracts of cattle. Br. J. Nut. 83:295–306.[Medline]
Wylie, M. J., T. W. White, W. C. Ellis, and J. H. Matis. 1990. The flow of undigested corn residues through the gastrointestinal tract of cattle. J. Anim. Sci. 68:3843–3852.[Abstract]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and expected for models of dose site = sample site (---) and dose site 