The need to develop an in vitro method to simulate ruminal digestion of protein foods has long been recognized. An in vitro method must be more rapid, consistent and convenient than in sacco methods. Purified enzymes have been examined in the past to estimate in vitro degradability of protein foods (Poos-Floyd et al., 1985; Aufrere et al., 1991; Luchini et al., 1996) or their fractions (Chaudhry and Webster, 1994). However, the selection of an appropriate enzyme for a range of foodstuffs remains to be realized. This study examined the potential of two enzymes and a centrifuged rumen fluid (CRF) to estimate proteolysis of food proteins.

Crude protein concentration and energy value of cereal crop residues and poor quality grasses are below animal requirements during much of the year and browses have been incorporated in the feeding systems to improve the nutritional status of ruminants. Leaves and fruits of leguminous browses can be used as protein supplements for ruminants and are commonly used as foods in many agricultural systems. The new systems of protein evaluation (Agricultural and Food Research Council, 1992; Tamminga et al., 1994) partition food nitrogen into the amount degraded in the rumen and that which escapes ruminal degradation. Protein available for absorption post-ruminally is influenced by the amount of food nitrogen that is resistant to ruminal degradation plus microbial protein (Brown and Pitman, 1991). Several studies have been carried out to assess the digestibility of ruminal escape protein (Hvelplund, 1985; Frydrych, 1992), and the digestibility varies considerably across foods (Krishnamoorthy et al., 1982). Values between 0·80 and 0·85 are adopted for the true digestibility of protein from temperate foods and from microbial amino acids (van Bruchem et ah, 1989). However, no data are available on intestinal digestibility of browses although they have been recognized as good protein supplements. The mobile nylon bag (MNB) method is useful for assessing the digestibility of rumen undegradable food protein in the intestine (de Boer et ah, 1987). Since the MNB is labourious, it was desirable to assess the accuracy with which an in vitro (pepsin/pancreatin) procedure could predict results from the former. Therefore, the objective of this study was to determine apparent digestibility of browse rumen undegradable protein using MNB method and pepsin/pancreatin in vitro technique.

The total amount of dietary protein available for absorption is dependent on the flow of dietary nitrogen (N) to the duodenum and its intestinal digestibility. Variation in intestinal digestion among protein supplements has been reported in vivo (Stern et al., 1985) and in situ (Hvelplund, 1985). Obtaining estimates of protein digestion in the small intestine is expensive, labour intensive and requires the use of surgically prepared animals. The objective of this study was to develop a reliable in vitro technique to estimate intestinal digestion of proteins in ruminants.

Expressing the protein value of a food involves measurements of several of its characteristics. Many in vivo studies have shown, that the protein degradability in the rumen varies substantially both between and within foods and therefore estimation of protein degradability in the rumen is an important task in protein evaluation. The most common method used has been the in situ (in sacco, nylon bag) method but many in vitro methods have been introduced and are based on use of either buffer solubility, chemical methods, rumen fluid or enzymes. None of these in vitro methods has proven to be of general use. In further development of in vitro methods as well as the in situ method a major problem is lack of a set of samples with a ‘true’ in vivo degradability which can be used for calibration of alternative methods. Microbial protein synthesis in the rumen has to be related to food characteristics which can be analysed easily. In vitro methods which can predict organic matter digestibility in foods are available and can be used to predict microbial protein synthesis in the rumen. Intestinal digestibility of undegraded dietary protein varies substantially both between and within foods and easy methods to estimate intestinal digestibility are therefore essential. The mobile bag method is easy to use and seems to give reliable results on most foods but requires access to duodenal cannulated animals which prevents this method from being routine. Alternative in vitro methods have been developed but further research is required for validation of these methods on a wide range of foods before they can be accepted for general use.

Ruminant and ruminal responses to feeding the gram-positive ionophore, monensin, have been researched extensively over the past 20 years. A proportion of many such in vivo experiments have used a change-over design. In doing so, the researcher either paid no attention to or was reasonably confident that any possible carry-over effects would have dissipated. Evidence does exist which leads to an estimation of duration to maximum treatment effects but such comparable evidence surrounding the duration of carry-over to monensin treatment is not available.

An in vivo trial was proposed at the Centre for Dairy Research (CEDAR), to investigate the ruminal effects of feeding monensin to lactating dairy cows with 4-week periods of which 3 weeks was for change-over and adaptation. Therefore, before conducting such an expensive experiment, in terms of both time and money, an in vitro study using the rumen simulation technique (Rusitec) was carried out to determine if 3 weeks was considered adequate to eliminate carry-over effects when measuring volatile fatty acid (VFA) molar proportions.

In vitromethods for laboratory estimation of food degradation are important tools for nutritionists.These methods either measure substrate disappearance by quantifying incubation residues or record fermentation products such as microbial biomass, short-chain fatty acids (SCFA) or gas volume (Blümmel et al., 1997a). Recently, the surge of interest in the efficient utilization of roughage diets has caused an increase in the use of gas methods because of the possibility of estimating the extent and rate of degradation in one sample by time series measurements of the accumulating gas volume. We combined gas measurements with residue determinations (truly degraded substrate) or microbial mass determination after fermentation to study the effects of natural plant products, in particular tannins and saponins, on the availability and partitioning of nutrients.

Ruminant production is an important sector of livestock agriculture. However, compared with simple-stomached animals, the conversion of plant material to animal products is low in efficiency and gives rise to large amounts of waste to the environment. Improvement in animal performance in the past has been based largely on greater input into the production system, rather than on a better efficiency of food conversion. With the increasing focus on resource efficiency and environmental protection, one important feature of future production systems will be to maximize efficiency and minimize waste. To aid the development and management of such systems, rapid food evaluation techniques are desired. In this paper, we propose that the measurement of purine derivative nitrogen as a proportion of total nitrogen in the urine, could provide a rapid indication of the efficiency with which food nitrogen is converted to microbial protein.

Maintaining blood pH within a narrow range is the first priority of an animal compared with the other body functions (Erdman et al., 1982). Under normal forage feeding conditions, maintenance of rumen, blood and cellular pH within safe ranges poses no problem for the animal. However, the inclusion of high levels of readily fermentable carbohydrates is often necessary to meet the energy requirements of high producing ruminants. Unfortunately, these foods can result in low rumen pHs which can have adverse effects on animal health and performance. When saliva flow is inadequate to counteract excessive ruminal acidity, the use of buffers may be justified. The in vivo studies reviewed by Muller and Kilmer (1979) have shown that the responses to the addition of buffers to dairy cow rations has been variable and inconclusive.

Part of the reason for this uncertainty is because buffer requirements cannot be accurately predicted from the chemical composition of foods, since measurements such as buffering capacity (BC) take no account of changes during fermentation such as the production of volatile fatty acids (VFA), ammonia and the disappearance of the fibre which confers BC. An in vitro approach was adopted for this work since it is difficult to conduct in vivo experiments in this area, owing to the health risks to cows as well as difficulties in distinguishing diet effects because of the strategies employed by cows to overcome excess acidity (e.g. eating rate, salivation, rumination).

There are two major ways to model the microbial activity in the rumen. The first is the material or physical modelling with in vitro systems and the other is mathematical modelling with sets of equations.

The evaluation of the ability of in vitro devices is to mimic in vivo ruminal digestion needs to perform first a typology of the most known and published in vitro procedures. This typology is aimed at clustering them according to their basic function (batch or continuous), to their major targets (studies of food degradation, of microbial proteosynthesis, of volatile fatty acid (VFA) production and profile, of gas production and composition, of outflow influences, of interactions between microbe species … also of prediction of intake of food nutritive value), to their level of time integration (cumulative at a given time or kinetic values), to their actual validity to simulate one or several items of ruminal digestion. Another key of typology of in vitro procedures is also the nature of the active substance, microbes or enzymes.

Rumen mathematical modelling can be variously used in connection with in vitro and in vivo experimentations. At least three main targets can be pursued in this perspective:

To model in vivo ruminal digestion. This aspect already has a long history because the first actual mechanistic model of whole rumen digestion was that of Baldwin et al. (1970). Since this date around a tenth of whole rumen models were proposed. Their current limits are those of the knowledge on rumen digestion: proteosynthesis, VFA production, nutrient absorption, liquid and particle outflows. As far as we are aware, none of these published rumen models was checked against in vitro data. One reason could be that some of them were more or less built from in vitro data.

To model in vitro devices. Surprisingly no such specific model seemed to be published until now. In our opinion it is urgent to develop such models because they could be highly useful: to compare analogies and discrepancies between the various in vitro procedures; to try to build a unique model of in vitro ruminal digestion which could be parameterized according to the type of device; to elucidate intrinsic properties of micro-organisms.

To model both in vitro and in vivo digestion. This way is likely to be the most difficult but it is also potentially the most fruitful. Effectively the use of an intermediate mathematical model of the rumen seems to be a priori the best way to check the major areas of analogy and discrepancy between in vivo and in vitro events. Thus this approach could be an objective mean of validation of in vitro studies.

In protein evaluation systems for ruminants, the microbial protein supply is calculated from the amounts of rumen degradable organic matter and nitrogen (N) using empirical equations. A variable part of the rumen synthesized microbial protein does not reach the duodenum but is recycled within the rumen (review Firkins, 1996). Since energy is required for its re-synthesis and degraded microbial protein is subject to deamination, the efficiency of substrate conversion into microbial protein in the rumen is affected by microbial recycling. Rumen protozoa have a major impact upon this recycling through engulfment of micro-organisms and autolysis. In vitro, bacterial protein breakdown is proportionately reduced by some 0·9 upon removal of protozoa (Wallace and McPherson, 1987). Defaunation of the rumen increases the efficiency of microbial protein synthesis in vivo significantly (review Jouany et al., 1988).

Ruminant animal production faces numerous challenges and it seems that both economic and biological benefits will be derived from moving food characterization from simple energy and protein-based approaches to those which assess nutrient supply in some detail. In vitro systems will need to reflect this change and this paper considers in particular, the need for estimations of rumen volatile fatty acids and microbial protein supplies. Emphasis is placed on the possibility that in vitro techniques can be used to provide biochemical data which can themselves be used in mathematical models of wider processes. This paper also examines the need for in vitro techniques to reflect the microbial!animal response to the physical structure of foods and also the requirement for in vitro approaches which ask why a food has a certain value rather than simply what the value is. In vitro techniques also have a substantial role outside the digestive tract in predicting factors such as voluntary food intake and some aspects of tissue metabolism and some of these aspects are considered. Tor practical application in vitro techniques will need to provide value for money and be compatible as parameters in mathematical models to have an impact at farm level. In this regard physical in vitro techniques such as NIRS seem to have enormous potential.

Parameters from in vitro gas production techniques could have potential as predictors of dry-matter intake (DMI) and digestibility. Fermentation is usually carried out under conditions where nitrogen (N) is not limiting. Therefore where N supply is a constraint to intake and digestibility, prediction equations may be inaccurate. This study compared the use of N-free and N-rich media in an in vitro fermentation method (Theodorou et al., 1994) and studied the relationships between in vitro and in vivo parameters obtained using both media.

Previous work in animal nutrition has focused on single foods and assumed additivity during in vitro fermentation. In the tropics, farmers are likely to offer mixtures of foods, including tree fodders, which may not be simply additive in nutritional terms. There is little information about the nutritional interactions between tropical foods. The objective of this research was to test the existence of associative effects of mixtures of tropical fodder tree leaves.

In the tropics, fodder trees and shrubs are a very important source of nutrients, especially nitrogen. In vitro gas production methods used for food evaluation were originally developed for investigation of temperate forages and used a nitrogen-rich medium. Evaluation of fodder tree leaves in this medium may mask the effect of their nitrogen which it is important to understand. This trial studied the fermentation of a range of tropical fodder trees and shrubs in both nitrogen-rich and nitrogen-free media, in order to identify the main chemical entities contributing to gas production and the time within which such contributions were most important.

Bioassay methods to estimate the digestibility of forages for ruminants, such as the in vitro digestibility technique (Tilley and Terry, 1963), the nylon bag technique (Ørskov et al., 1980) and the gas production methods of Menke and Steingass (1988) and Theodorou et al. (1994), require rumen fistulated animals, either to provide a suitable in situ environment or to provide rumen liquor as a source of inoculum. Not only is establishing and maintaining fistulated animals expensive, but fistulation is an invasive technique which is increasingly discouraged on animal welfare grounds. There is therefore a need to find an alternative to rumen liquor as a source of micro-organisms for bioassays.

Although the Tilley and Terry (1963) technique is widely used, it is limited by being an end-point digestibility method. Ørskov et al. (1988) showed that intake of forages and their rate of digestion in the rumen are more correlated than intake and digestibility. Thus, since 1988, there has been much interest in determining rate of rumen degradability using the nylon bag technique (Huntington and Givens, 1995). However, as indicated earlier, this in vitro technique requires fistulated animals. Recently Sileshi et al. (1996) showed that the in vitro gas production technique of Theodorou et al. (1994), offers a possibility of assessing rate of rumen degradation.

The purpose of the present experiment was to compare rumen liquor and faeces as sources of inoculum in the gas technique of Theodorou et al. (1994).

The in vivo organic matter digestibility (OMD) in ruminants can be predicted with the gas test method by the amount of gas produced after 24-h fermentation (Menke et al., 1979). Mathematical models can also be used and can provide useful information concerning the kinetics of fermentation. In ruminants, various models have been used to fit fermentation profiles. The early models are based on first-order kinetics with a constant fractional rate of fermentation (Ørskov and McDonald, 1979; Khazaal et al., 1993). Recently, more sophisticated models have been used to describe gas production kinetics obtained from new automated gas production equipment (Theodorou et al., 1995; Cone et al., 1996). The aim of this study was to compare the relevance and the accuracy of five models for describing gas production kinetics and for predicting the organic matter digestibility (OMD) of forages in horses.

The gas test is commonly described as a simple, rapid and cheap technique for predicting the energy value of forages but it requires animals fitted with a canula. In horses, this method can be implemented with caecal fluid as an inoculum to predict the organic matter digestibility (OMD) of forages (Macheboeuf et al., 1998). In the present work, as bacterial content of faeces was known to be very high (Meyer et al., 1993), the use of faeces as a source of inoculum for predicting OMD of forages in horses was studied.

Gas production from the in vitro digestion of forage with buffered rumen fluid can be measured and used to determine forage digestibility and fermentation kinetics. Rumen micro-organisms ferment carbohydrate to gases (CO2 and CH4) and volatile fatty acids (VFA). The VFA produced also cause CO2 to be released from the C02-bicarbonate buffer. Theodorou et al. (1994) introduced the principle of measuring gas production by pressure increase using an electronic pressure transducer and sealed gas-tight culture bottles. Gases accumulate in the head space of the culture bottles as fermentation proceeds. The gas is measured and then released at regular intervals throughout the fermentation. This procedure was automated (Davies et al., 1995). The automated pressure evaluation system (APES) has advantages over the manual pressure transducer technique (Theodorou et al., 1994) in that it is less labour intensive and has been shown to be more sensitive to food characteristics (Davies et al., 1995). The APES, used in this work, has been improved to include new switches and a filtering system. It has been used here to determine the fermentation characteristics of various ruminant foods: ryegrass, wheat grain and ryegrass silage.

Pressure transducer technology to measure gas production from microbial ecosystems has been utilized in a number of ways but predominantly for food evaluation. The approach also has considerable potential to increase our understanding of, and ability to manipulate, the rumen microbial ecosystem, but most research to date has concentrated on measurement of total gas production and not composition. The aim of this study was to extend the scope of the gas production technique to the quantitation of component gases, whilst investigating ruminal gas production in the presence and absence of methanogens.

Inter- and even intra-species variability exists in the ability of rumen inocula to digest forages using in vitro techniques. Differences in fermentation patterns between animal species may be due to the presence of different species within the microbial populations and/or to the fermentative capacity of the microflora (Sunvold et al., 1995). Several authors have reported that the host species affects the rumen microflora which therefore influences the in vitro activity of rumen fluid, even if two host species are given the same diet. Grant et al. (1974) concluded that after a sufficiently long adaptation period to the diet, there was no longer any significant difference between the rumen fluid from buffalo and cows. However, the results obtained by other authors (di Leila et al. 1995; Puppo et al., 1995) are in disagreement with these findings. The in vitro cumulative gas production technique (Theodorou et al., 1994), is increasingly being used as a method of food evaluation, by comparing the fermentability of different foods. One of the greatest sources of variation for this kind of method is due to differences in inocula. An experiment was carried out to determine whether rumen fluid from buffalo and cows given a similar diet, led to similar results for the evaluation of the same foods.