| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ANIMAL NUTRITION |
,
* Institute of Grassland and Environmental Research, Plas Gogerddan, Aberystwyth, Ceredigion SY23 3EB, U.K.;
and
Department of Animal and Poultry Science, University of Guelph, Guelph, Ontario N1G 2W1, Canada;
and
The University of Reading, School of Agriculture, Policy and Development, Whiteknights, Reading RG6 6AR, U.K.;
and
Fundacao Ezequiel Dias 80, CEP 30550, Belo Horizonte MG, Brazil;
and
¶ WIAS Animal Nutrition Group, Wageningen University, 6709 PG Wageningen, The Netherlands; and
and
# Departamento de Producción Animal, Universidad de León, E-24007 León, Spain
Abstract
A method is proposed to determine the extent of degradation in the rumen involving a two-stage mathematical modeling process. In the first stage, a statistical model shifts (or maps) the gas accumulation profile obtained using a fecal inoculum to a ruminal gas profile. Then, a kinetic model determines the extent of degradation in the rumen from the shifted profile. The kinetic model is presented as a generalized mathematical function, allowing any one of a number of alternative equation forms to be selected. This method might allow the gas production technique to become an approach for determining extent of degradation in the rumen, decreasing the need for surgically modified animals while still maintaining the link with the animal. Further research is needed before the proposed methodology can be used as a standard method across a range of feeds.
Key Words: Extent of Degradation • Fecal Inoculum • Gas Production • Ruminal Inoculum
Introduction
In vitro techniques have been used extensively in feed evaluation and in studies of ruminal fermentation. A sample of feed is incubated in batch cultures, measuring substrate disappearance (in vitro digestibility) and/or end-product accumulation (fermentation gas) either at an endpoint or at a sequence of different time points. Mixed ruminal microorganisms are the inoculum of choice to recreate ruminal conditions. Ruminal liquor, even when diluted, provides sufficient microbes to achieve rapid degradation of feedstuff in vitro. An alternative to ruminal liquor as the inoculum, not requiring surgical intervention while maintaining the link to the animal, is fecal matter (El Shaer et al., 1987
). The fermentation process seems to be similar in cultures of ruminal or fecal microorganisms, as the pattern of end-product yield is similar when the same substrate is incubated (El-Meadaway et al., 1998). However, low levels of surviving microbes mean that inocula derived from fecal matter will exhibit reduced degrading potency. This will be expressed as a longer lag phase and slower degradation rate at the outset (Mauricio et al., 2001). Eventual cumulative gas production will be a little less than what is expected from ruminal liquor because not all the microbial species survive postruminal processes (El-Meadaway et al., 1998). Thus, mathematical adjustments will be required to convert or translate the degradation profiles produced from fecal inoculum to those obtained when the inoculum originates from ruminal liquor.
A less invasive method is proposed for determining the extent of degradation in the rumen based on the gas production technique and mathematical modeling. The objective of the study reported herein is to correct mathematically gas production profiles obtained from fecal matter inoculum to look like the reference profiles when using ruminal liquor. Preliminary reports of this work have been given in France et al. (2000a) and Crompton et al. (2001).
Experimental Procedures
The data used in this study are those reported by Mauricio et al. (2001) from two experiments in which matched and simultaneous in vitro gas production were measured with inoculum derived either from ruminal liquor or from fecal matter collected from the same donor animals. Data from Exp. 1 corresponded to 12 forages of differing in vivo OM digestibilities (range 548 to 807 g/kg), namely ammonia-treated wheat straw, hay (field cured, Lolium perenne), and 10 artificially dried grasses (Lolium perenne) cut at different growth stages. The NDF and ADF composition of these forages ranged from 488 to 764 and 239 to 362 g/kg of DM, respectively. In vitro gas profiles from Exp. 2 were for maize silages prepared from seven botanical fractions, defined as whole plant, stover [whole plant less ear (cob plus grain)], leaf, lower stem (maize fraction collected below the third node), middle stem (maize fraction collected between the third and sixth node), upper stem (maize fraction collected between the sixth node up to the ear), and husks (leaves covering the ear). The seven fractions were ensiled with and without an enzyme mixture (Enzyme C, Finnfeeds International, Marlborough, U.K.) applied at a level of 0.2 mL/kg fresh forage. The pH values and DM contents of these 14 silages ranged from 3.9 to 4.3 and from 191 to 476 g/kg, respectively. The NDF and ADF concentrations in the silages ranged from 412 to 693 and 210 to 365 g/kg of DM.
The in vitro gas production technique used in the study by Mauricio et al. (2001) was as described in Theodorou et al. (1994). One gram of substrate DM, 10 mL of inoculum, and 90 mL of medium were placed in each bottle. Inocula were prepared from ruminal and fecal matter as described by Mauricio et al. (2001). Ruminal liquor and fecal samples were obtained before the morning feed (0730) from two Friesian dairy cows in early lactation offered the same diet (9.4 kg of DM grass silage and 9 kg of DM concentrate daily). Gas readings were taken at 3, 6, 9, 12, 15, 21, 27, 33, 39, 48, 60, 72, and 96 h after inoculation. Initial readings were taken at 3-h intervals because of the rapid rate of gas production and the time required to take manual gas readings. Gas volumes obtained were corrected for the quantity of DM incubated and gas released from controls. For each time of reading, a mean value obtained from four bottles was used to generate gas production profiles. Unfortunately during data capture, the between-bottles variation was not retained. However, as a pilot study using the same gas production method, a repeatability and reproducibility experiment was conducted (results not published) with samples from 10 forage feeds with five replicate bottles and over five different gas production runs. There were 13 sampling times between 3 and 72 h of incubation. The percentage of the standard error of mean volume declined from 3.1% at 3 h to 1.0% at 72 h, with overall mean of 1.7%.
Results
Gas Production Profiles
The matched study by Mauricio et al. (2001) with inoculum derived from ruminal liquor or fecal matter has generated paired gas production profiles for the 12 grasses and 14 maize silages, as shown in Figures 1 and 2. These paired profiles differ in terms of lag time, and rise to and height of plateau, but their shapes show similarity over time.
|
|
Step 1.
The similarity between the gas profiles observed with different inocula suggests that the gas production profiles obtained using fecal matter can be shifted to reproduce those obtained using ruminal liquor. Plots of gas accumulation for ruminal liquor (y-axis) against gas accumulation for fecal matter (x-axis) at matched incubation times for each forage used in Exp. 1 and 2 are shown in Figures 3 and 4, respectively, illustrating a strong piecewise linear relationship. A linear spline (two-phase) model with a single unknown break point (node) was therefore fitted to each plot using the nonlinear regression package MLP (Ross, 1980), generating a break point, Bf (milliliters of gas accumulation per gram of substrate DM incubated for fecal matter), for each forage (Figures 3 and 4). Specific to the estimate of the break point, the slope and intercept of a straight line describing each phase were estimated simultaneously (Ross, 1990).
|
|
.
|
, 1990), and it is assumed that this area equals that of the triangle LVP. The base LP of the right-angled triangle LVP can now be calculated by equating this area to the formula (0.5 x LP x VP). The distance along the time axis from the origin to point L is the graphical estimate of the lag time. The choice of 0.4 times the maximum gas volume was based on rigorous testing, which involved estimating the lag graphically using different proportions of maximum gas volume and comparing these values with the best estimate of the lag obtained from applying the various models described by France et al. (2000b) and Dhanoa et al. (2000). Graphical estimates of lag were correlated with model estimates giving r2 values of 0.636 and 0.893 for Exp. 1 and 2, respectively. The proportion 0.4 may possibly vary with the gas production system in use or the frequency of gas sampling.
Step 3.
Three datasets were used to construct the statistical model. Dataset 1 comprised 10 grasses (i.e., the grasses from Exp. 1 but excluding the ammonia-treated straw and dried grass 9). Set 2 comprised 13 maize silages (i.e., the silages from Exp. 2 but excluding whole plant without enzyme mixture). Set 3 comprised all of the forages in Sets 1 and 2 plus the ammonia-treated straw (i.e., a total of 24 forages). Dried grass 9 and the whole plant maize silage were excluded randomly from the data sets so that they could be used subsequently for independent evaluation of the proposed method.
Figure 6 shows the plot of Bf against Tf (estimated using the graphical approach) and the line of best fit for Dataset 3. The values of gas accumulation for fecal matter at the break point, Bf (mL/g DM), (depicted in Figures 3
and 4) correlated with the estimates of lag time, Tf (h), obtained from the gas production profiles for fecal matter (depicted in Figures 1 and 2). The regression equations obtained for Datasets 1, 2, and 3, respectively, are:
|
 | [1a] |
 | [1b] |
 | [1c] |
These regression lines were derived using ordinary least squares (OLS) and verified using the Bootstrap method, which is a method for providing estimates with reduced bias (Efron and Tibshirani, 1993).
Step 4.
Having predicted break point Bf, equations to estimate gas accumulation for ruminal liquor from gas accumulation for fecal matter were fitted by linear regression. Points below their respective break point in Dataset 3 (these points are depicted in Figures 3 and 4) were pooled to produce Figure 7, which also shows the line of best fit derived using OLS and verified with the Bootstrap method. The regression lines obtained for each of the three data sets are:
|
 | [2a] |
 | [2b] |
 | [2c] |
The variables Vr and Vf (both mL/g of DM) denote gas accumulation for ruminal liquor and fecal matter, respectively. From an inspection of Figure 7, it is apparent that some data points at the top end cause clockwise rotation of the OLS line. If such data points are believed to exert undue leverage, then the fault can be corrected using nonparametric Theil regression (Sprent, 1993; Dhanoa, 1998).
Similarly, all the points above their respective break point in Dataset 3 (again these points are depicted in Figures 3 and 4) were pooled to produce Figure 8, which also shows the line of best fit. The regression lines for the three datasets are:
|
 | [3a] |
 | [3b] |
 | [3c] |
In summary, a gas production profile obtained using fecal matter as the inoculum can be shifted to a profile from ruminal liquor by estimating its lag time Tf and applying Eq. [1], [2], and [3] to adjust each observation. The lag time can be determined graphically or using a lagged model.
Step 5.
Finally, OLS regression equations were derived to relate the residue remaining in the flask after 96 h of incubation using ruminal liquor as the inoculum, Ur (g of DM), to that remaining using fecal matter, Uf (g of DM). The parameter Ur is required for subsequent calculation of extent of degradation in the rumen. The equations derived for each dataset were:
 | [4a] |
 | [4b] |
 | [4c] |
These equations were also verified by the Bootstrap method. The effect of measurement errors in the predictor variable Uf was also examined, using an alternative regression method called least rectangles regression (Sokal and Rohlf, 1994). This examination confirmed that the OLS regression equations are acceptable, mainly due to the high correlation between Ur and Uf. However, alternative regression methods may be needed when the correlation is low (r < 0.7).
Kinetic Model
The substrate incubated in a flask is assumed to comprise two components, a potentially degradable component S and an undegradable component U (both g of DM). The quantity of S remaining in the flask at time t (h) after the start of incubation can be represented by the general expression:
 | [5] |
where S0 (g of DM) is a positive scalar denoting the portion of the incubated substrate that is potentially degraded over time, 
is a positive monotonically increasing function within the range [0,1] with an asymptote at = 1, and T (h) is the lag time. The fractional rate of degradation µ(h-1) is given by:
 | [6] |
It follows from Eq. [5] that cumulative gas production to time t, V (mL), can be written:
 | [7] |
where Y (mL/g DM) is a constant yield factor. Equation [7] gives a generalized mathematical function for describing cumulative gas production profiles, allowing any one of a number of alternative equation forms to be selected for fitting.
The fractional rate of degradation, providing it represents events in the rumen, can be combined with rate of passage to calculate the extent of ruminal degradation (effective degradability) of the substrate. If S is the amount of potentially degradable substrate remaining in the rumen and subjected to both passage and degradation, the rate of disappearance of S is given by:
 | [8a, b] |
where k (h-1) is the fractional rate of passage. This parameter can be obtained using any available model for passage kinetics (France et al., 1985).
The solutions of these differential equations are:
By integration of Eq. [8a]:
 |
By integration of Eq. [8b]:
 |
which gives:
 | [9] |
Using Eq. [6] and [9] and given a value for k, the extent of degradation in the rumen, E (g of DM degraded/g of DM ingested), is given by the equation:
 |
where:
 |
as is zero when t equals T. Therefore:
 | [10] |
where 
e-ktdt can be calculated by numerical integration if no analytical solution exists.
In summary, we have derived an expression for extent of degradation in the rumen (Eq. [10]), which is applicable to any model for describing gas production profiles expressed in the form of Eq. [7].
Model Evaluation
To test the method proposed in this study, dried grass 9 from Exp. 1 and whole plant maize silage from Exp. 2 (see Figures 1 and 2) were used. These two forages were excluded randomly from each of the three data sets used to develop the method so that they could be used subsequently for independent evaluation. The gas production profile, obtained with fecal matter, for dried grass 9 was shifted to a profile from ruminal liquor according to both the equations for grass (Eq. [1a] - [3a]) and the equations for forage (Eq. [1c] - [3c]), producing two estimates of the profile for ruminal liquor. Similarly, the profile from fecal matter for whole-plant maize silage was shifted to a profile from ruminal liquor according to both the equations for maize silage (Eq. [1b] - [3b]) and the equations for forage (Eq. [1c] - [3c]).
To test the ability of the shifted fecal profiles to mimic the actual profiles obtained with ruminal liquor, Lin’s concordance coefficient 
c (sample equivalent rc; Lin, 1989; Dhanoa et al., 1999) and intraclass correlation 2 (sample equivalent r2; McGraw and Wong, 1996) were used. These indices of reproducibility assess the linear relationship between two variables under the constraints that the intercept is 0 and the slope is 1, unlike simple correlations where the relationship is quantified without these constraints. The calculated values for the two chosen forages are presented in Tables 1 and 2 (simple correlation r is given for reference only as it can be misleading when testing reproducibility). Given that the values of rc and r2 are very similar and close to unity (perfect match), similarity between the actual and estimated profiles is established.
|
|
, were analyzed using the model of France et al. (1993) (denoted as F in Table 3 and 4) and the model of Ørskov and McDonald (1979) (denoted as ØM in Table 3 and 4), namely:
|
|
|
 |
where b, c and T are parameters to be estimated (along with the product coefficient Y x S0). The model of Ørskov and McDonald (1979) for in situ studies, as refined by McDonald (1981), is a special case of that of France et al. (1993) (c = 0). The parameter estimates obtained from fitting the models are presented in Table 3. Derived quantities (the integral e-ktdt and extent of degradation in the rumen) obtained using Ur predicted from Uf and two assumed values of rate of passage k (h-1) are given in Table 4. There were negligible differences in the estimated value of E for the maize silage, whereas small differences were detected for the dried grass, probably attributable to overestimation of Ur and T using the gas profiles observed with fecal matter inoculum.
The in vivo OM digestibility of the grass hay and the 10 dried grass samples used in the gas production study were determined in four sheep fed at maintenance as described by Givens and Moss (1994) and reported by Maurcio (1999). The in vivo OM digestibility and the calculated values for the extent of degradation in the rumen, using two assumed rates of passage, are presented in Table 5. In addition, Table 5 includes the correlation coefficients between in vivo OM digestibility and the calculated extents of degradation using a simple correlation and Spearman’s rank correlation. The data indicate that at both assumed rates of passage, in vivo OM digestibility was correlated with extent of degradation in the rumen, calculated from both the actual and estimated gas profiles. Furthermore, the calculated extents also have a ranking order similar to that of in vivo OM digestibility, as shown by the Spearman’s rank correlation coefficient (see Table 5).
|
Recent work on in vitro gas production with inocula derived from fecal matter suggests that it might be possible to move away from ruminal liquor sampling (Akhter et al., 1999; Mauricio et al., 2001). In vitro gas production based on ruminal liquor is limited by the need for donor animals and facilities for surgical intervention, whereas gas production based on fecal matter has no such limitations. Under this system, research is not restricted to housed herbivores, and animals can be studied in their natural environment behaving as their biology dictates. Furthermore, it creates opportunities to study exotic species of herbivores in the wild.
The study by Mauricio et al. (2001) comprising Exp. 1 and 2 was set up to compare gas production results obtained in vitro using ruminal liquor with corresponding results obtained using fecal matter from the same donor animals. Usually, a simple regression approach is applied to 1) establish the statistical correlation between gas production profiles obtained using fecal matter and ruminal liquor as the inoculum (Mauricio et al., 2001) and 2) predict ruminal degradability or whole tract digestibility from parameters estimated from fecal profiles (Aiple et al., 1992; Akhter et al., 1999; Borba et al., 2001). The determination of the extent of degradation in the rumen is based on more mechanistic models developed from the assumption that degradation kinetics in cultures of mixed ruminal microorganisms reproduce in vitro the kinetics in the rumen. Thus, parameters estimated from gas profiles produced using fecal matter as the inoculum should not be used directly to predict extent of degradation in the rumen. It seems a more rational approach to shift the fecal matter profiles to simulate or estimate one based on ruminal liquor that can then be used to estimate the extent of degradation applying the mechanistic model. The model developed herein, although including an empirical element in the first stage, is based on the biological link between parameters rather than on the statistical relationship between variables as in most purely empirical models. The kinetic model allows for the estimation of the extent of degradation in the rumen from in vitro data using a mechanistic approach.
The close relationship between data obtained with ruminal or fecal inocula has been reported widely (Akhter et al., 1999; Borba et al., 2001; Mauricio et al., 2001). The method proposed in the present work is based on this similarity and successfully converts gas production profiles obtained using fecal matter as the inoculum to profiles that resemble those obtained using ruminal liquor. However, more fermentation gas is produced with a ruminal than with a fecal inoculum and, although the ranking of feeds using one or the other is similar, within each profile, there is a biphasic relationship in gas volume recorded from cultures of ruminal or fecal microorganisms. The plots of ruminal gas volumes against fecal gas volumes (Figures 3 and 4) reveal that both phases are linearly related, but with significantly different slopes—greater than unity below the break point and slightly lower than unity above the break point. Thus, the difference between ruminal and fecal gas volumes increases during early incubation and is maintained or even reduced slightly during prolonged incubation. This confirms that fecal matter has reduced fermentation activity compared to ruminal liquor, resulting in lower gas volume, longer lag time, slower fermentation rate and lower OM disappearance (Mauricio et al., 2001).
Fecal microorganisms originate from the hindgut where fermentation activity is lower than in the rumen for a number of reasons. Firstly, total and cellulolytic bacterial populations and the diversity of bacterial species are higher in the rumen than in the hindgut or in feces (Kern et al., 1974; El-Meadaway et al., 1998; Omed et al., 2000). Secondly, fecal suspensions may contain some anaerobic fungi (Davies et al., 1993) but lack protozoa (Hobson, 1971). Thirdly, the microorganisms in feces are perhaps in a state of low metabolic activity in contrast with the rumen where they show an actively growing state (Mauricio et al., 2001). Finally, ruminal liquor supplies some trace elements, micronutrients, and unknown factors deemed essential for microbial growth that may not be present in fecal suspensions (Omed et al., 2000). Therefore, inoculum type, size, and activity have a profound effect on microbial growth and substrate degradation in batch cultures, and thus on gas production.
Cultures inoculated with fecal matter need longer to achieve their degradation potential than with ruminal liquor, but once the microbial numbers reach a given threshold, the rate of gas production is similar in both cases. Hidayat et al. (1993) suggested that once the maximal degradation rate is achieved, the addition of more microorganisms fails to stimulate further digestion. In our case, the threshold would be represented by the break point for the plots of ruminal against fecal gas, as from that point the slope is lower than unity indicating that the increase in gas production between two incubation times is relatively larger with the fecal than with the ruminal inoculum. If the incubation time were indefinitely long, total gas production would be similar in volume regardless of the source of inoculum (El-Meadaway et al., 1998; Omed et al., 2000). Thus, our interest in estimating break point on the basis of the likely biological link between break point and an unbiased predictor not derived from fitting any model to avoid dependency on model choice. The work reported suggests that the break points generated may have some biological and/or chemical basis, but exactly what microbial and substrate attributes are important requires further research.
Some limitations are inherent in any in vitro technique, especially one tracking ruminal degradation kinetics (Lopez et al., 2000). Whereas techniques using ruminal liquor have been studied extensively, much less is known about the conditions required to achieve optimal microbial growth and substrate degradation when using a fecal inoculum for the culture. This work is required to improve the repeatability and reproducibility of the fecal technique. Gas production and substrate fermentation in cultures inoculated with fecal matter are affected by factors such as composition of the incubation medium (buffer, addition of a nitrogen supplement), feces:buffer ratio, inoculum:substrate ratio, and inoculum preparation (Omed et al., 2000). Fibrolytic enzymes in fecal suspensions are provided by particle-bound microorganisms that need to be detached and recovered in the suspension during the preparation process. It is noteworthy that the activity of fecal matter inoculum is affected by dietary factors to a lesser extent than that of ruminal liquor (Omed et al., 2000).
The set of dried grasses and maize silages on which the method is based is small and further research is required to provide matched profiles for these and a range of other forages. Nevertheless, it is encouraging to observe the apparent success of the proposed method in relation to dried grass and maize silage (Table 4). Earlier studies have reported the existence of outliers in the relationship between in vitro gas production or digestibility measured with fecal matter and that measured with ruminal inocula (Mauricio et al., 2001), or the lack of a statistically significant correlation for some types of feeds such as silages (Borba et al., 2001) and poor quality roughages (El-Meadaway et al., 1998). The method developed herein has revealed that the relationship between gas volumes obtained with either ruminal or fecal inocula can be dependent on the type of feed used, requiring different predictive equations for each group of feeds.
This method is being made available on the Web (http://www.aps.uoguelph.ca).
Implications
A less invasive method for determining the extent of degradation in the rumen, which involved developing both a statistical model and a kinetic model, is proposed. The dataset used to produce the statistical model, comprising data on dried grasses and maize silages, is relatively small and further experiments are required to provide matched profiles for these and a range of other forages. It is possible that feeds with more complex compositions may not respond well to this approach because the population of fecal microbes may be sufficiently different from that found in rumen. The kinetic model is presented as a generalized mathematical function, allowing any one of a number of alternative equation forms to be selected. These advances might allow the use of fecal inoculum in the gas production technique to become a method for determining extent of degradation in the rumen, decreasing the need for surgically modified animals, while still maintaining the link with the animal.
Footnotes
1 We are grateful to DEFRA for financial support under contract LS3602. Support for J. France from the Spanish Secretaría de Estado de Educación y Universidades del Ministerio de Educación, Cultura y Deporte and the European Social Fund (Ayuda para estancias de profesores, investigadores, doctores y tecnólogos extranjeros en España Ref. SAB2000-0112) is gratefully acknowledged. RMM thanks the Conselho Nacional de Desenvolvimento Cientifico e Technologico (CNPq), Brazil for financial assistance. 
2 Correspondence—fax: 519-836-9873; e-mail: jfrance{at}uoguelph.ca.
Received for publication May 5, 2003. Accepted for publication November 10, 2003.
Literature Cited
Aiple, K. P., H. Steingass, and K. H. Menke. 1992. Suitability of a buffered faecal suspension as the inoculum in the Hohenheim gas test. 1. Modification of the method and its ability in the prediction of organic matter digestibility and metabolizable energy content of ruminant feeds compared with rumen liquor as inoculum.J. Anim. Physiol. Anim. Nutr. 67:57–66.
Akhter, S., E. Owen, M. K. Theodorou, E. A. Butler, and D. J. Minson. 1999. Bovine faeces as a source of microorganisms for the in vitro digestibility of forages. Grass Forage Sci. 54:219–226.
Borba, A. E. S., P. J. A. Correia, J. M. M. Fernandes, and A. F. R. S. Borba. 2001. Comparison of three sources of inocula for predicting apparent digestibility of ruminant feedstuffs. Anim. Res. 50:265–273.
Crompton, L. A., J. France, E. Kebreab, J. A. N. Mills, M. S. Dhanoa, S. López, and J. Dijkstra. 2001. A non-invasive approach to determining extent of degradation in the rumen. Page 26 in Proc. Br. Soc. Anim. Sci., Br. Soc. Anim. Sci., Edinburgh, Scotland.
Davies, D. R., M. K. Theodorou, M. I. G. Lawrence, and A. P. J. Trinci. 1993. Distribution of anaerobic fungi in the digestive tract of cattle and their survival in feces. J. Gen. Microbiol. 139:1395–1400.
Dhanoa, M. S. 1998. A procedure for Theil’s regression method. Genstat Newsletter 34:21–26.
Dhanoa, M. S., S. J. Lister, J. France, and R. J. Barnes. 1999. Use of mean square prediction error analysis and reproducibility measures to study near infrared calibration equation performance. J. Near Infrared Spectrosc. 7:133–143.
Dhanoa, M. S., S. Lopez, J. Dijkstra, D. R. Davies, R. S. Sanderson, B. A. Williams, Z. Sileshi, and J. France. 2000. Estimating the extent of degradation of ruminant feeds from a description of their gas production profiles observed in vitro: comparison of models. Br. J. Nutr. 83:131–142.[Medline]
Efron, B., and R. J. Tibshirani. 1993. An Introduction to the Bootstrap. Chapman and Hall, London, U.K.
El-Meadaway, A., Z. Mir, P. S. Mir, M. S. Zaman, and L. J. Yanke. 1998. Relative efficacy of inocula from rumen liquor or fecal solution for determining in vitro digestibility and gas production. Can. J. Anim. Sci. 78:673–679.
El Shaer, H. M., H. M. Omed, A. G. Chamberlain, and R. F. E. Axford. 1987. Use of faecal organisms from sheep for the in vitro determination of digestibility. J. Agric. Sci. (Cambridge) 109:257–259.
France, J., J. H. M. 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., M. S. Dhanoa, M. K. Theodorou, S. J. Lister, D. R. Davies, and D. Isac. 1993. A model to interpret gas accumulation profiles with in vitro degradation of ruminant feeds. J. Theor. Biol. 163:99–111.
France, J., M. S. Dhanoa, L. A. Crompton, S. Lopez, J. Dijkstra, R. M. Mauricio, E. Kebreab, and E. Owen. 2000a. A non-invasive approach to determining extent of degradation in the rumen. Pages 33–41 in Gas Production: Fermentation Kinetics for Feed Evaluation and to Assess Microbial Activity. Proc. EAAP Satellite Symp. on Gas Prod., Wageningen, The Netherlands. Br. Soc. Anim. Sci., Edinburgh, Scotland.
France, J., J. Dijkstra, M. S. Dhanoa, S. Lopez, and A. Bannink. 2000b. Estimating the extent of degradation of ruminant feeds from a description of their gas production profiles observed in vitro: Derivation of models and other mathematical considerations. Br. J. Nutr. 83:131–142.
Givens, D. I., and A. R. Moss. 1994. Effect of breed, age and bodyweight of sheep on the measurement of apparent digestibility of dried grass. Anim. Feed Sci. Technol. 46:155–162.
Hidayat, B., K. Hillman, C. J. Newbold, and C. S. Stewart. 1993. The contributions of bacteria and protozoa to ruminal forage fermentation in vitro, as determined by microbial gas production. Anim. Feed Sci. Technol. 42:193–208.
Hobson, P. N. 1971. Rumen microorganisms. Proc. Ind. Microbiol. 9:42–77.
Kern, D. L., L. L. Slyter, E. C. Leffel, J. M. Weaver, and R. R. Oltjen. 1974. Ponies vs. steers: Microbial and chemical characteristics of intestinal digesta.J. Anim. Sci. 38:559–564.
Lin, L. I.-K. 1989. A concordance correlation coefficient to evaluate reproducibility. Biometrics 45:255–268.[Medline]
Lopez, S., J. France, J. Dijkstra, and M. S. Dhanoa. 2000. Models to interpret degradation profiles obtained from in vitro and in situ incubation of ruminant feeds. Pages 73–86 in Modelling Nutrient Utilization in Farm Animals. J. P. McNamara, J. France, and D. E. Beever, ed. CABI Publishing, Wallingford, U.K.
Mauricio, R. M. 1999. Comparison of bovine rumen liquor and bovine feces as sources of microorganisms for an in vitro gas production technique for evaluating forages. Ph.D. Thesis, The Univ. of Reading, Reading, U.K.
Mauricio, R. M., E. Owen, F. L. Mould, D. I. Givens, M. K. Theodorou, J. France, D. R. Davies, and M. S. Dhanoa. 2001. Comparison of bovine rumen liquor and bovine faeces as inoculum for an in vitro gas production technique for evaluating forages. Anim. Feed Sci. Technol. 89:33–48.
McDonald, I. 1981. A revised model for the estimation of protein degradability in the rumen. J. Agric. Sci. (Cambridge) 96:251–252.
McGraw, K. O., and S. P. Wong. 1996. Forming inferences about some intraclass correlation coefficients. Psychol. Methods 1:30–46.
Omed, H. M., D. K. Lovett, and R. F. E. Axford. 2000. Faeces as a source of microbial enzymes for estimating digestibility. Pages 135–154 Forage Evaluation in Ruminant Nutrition. D. I. Givens, E. Owen, R. F. E. Axford, and H. M. Omed, ed. CABI Publishing, Wallingford, U.K.
Ørskov, E. R., and I. McDonald. 1979. The estimation of protein degradability in the rumen from incubation measurements weighted according to rates of passage. J. Agric. Sci. (Cambridge) 92:499–503.
Ross, G. J. S. 1980. Maximum Likelihood Program (A Manual). Rothamsted Exp Stn., Harpenden, U.K.
Ross, G. J. S. 1990. Nonlinear Estimation. Springer-Verlag, New York.
Sokal, R. R., and F. J. Rohlf. 1994. Biometry: The Principles and Practice of Statistics in Biological Research. 3rd ed. W. H. Freeman and Co., New York.
Sprent, P. 1993. Applied Nonparametric Statistical Methods. 2nd ed. Chapman and Hall Ltd., London, U.K.
Theodorou, M. K., B. A. Williams, M. S. Dhanoa, A. B. McAllan, and J. France. 1994. A simple gas production method using a pressure transducer to determine the fermentation kinetics of ruminant feeds. Anim. Feed Sci. Technol. 48:185–197.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or fecal matter (
) as the inoculum. The solid lines show the equation of France et al. (1993)
).
