### Important! This text corresponds to the INRA-AFZ 2002 Tables and has not been updated yet.

*Daniel Sauvant, Patrick Chapoutot, Jean-Louis Peyraud, François Meschy and Brigitte Doreau*^{U}

### Digestibility and energy value

#### Source of data and harmonisation of digestibility values

The digestibility and energy values for ruminants presented in the tables have been established using a database of digestibility measurements obtained *in vivo* on more than 300 feed samples studied by INRA (France) and by the Rowett Research Institute (United Kingdom). Since the digestibilities of organic matter (OMd), dry matter (DMd) or energy (Ed) were not available simultaneously for every feed sample, it was necessary to work on a unique parameter: OMd was present on 89% of the samples and the missing OMd values were calculated by regression equations using Ed or DMd as predictors. These equations integrated additional analytical characteristics when they were statistically significant.

#### Calculation of organic matter digestibility values for feed materials

With OMd values available for each feed material, we studied prediction models based on cell wall parameters (crude fibre, NDF, ADF) for this digestibility. Many of the resulting equations have crude fibre as the main predictor, as this parameter was the most frequent cell wall criterion in the database. Regressions integrating discriminating tests between groups of feed materials within the same relationship were calculated by variance-covariance analysis. Interactions between the predictor and the groups were also tested in order to identify the possible effects of the type or family of feed materials on the regression slope. Therefore, six major prediction models concerning the following groups of ingredients were calculated.

##### Cereal grains and cereal by-products

OMd = 95.81 – 1.911 CF + α (n = 124; r = 0.93; RSD = 3.7)

Where α = -2.54 all feed materials except maize by-products

α = +2.54 maize by-products

##### Full fat rapeseed, full fat sunflower, full fat cottonseed, copra meal, palm kernel meal and cottonseed meal

OMd = 97.51 – 1.498 CF (n = 29; r = 0.79; RSD = 6.5)

##### Legume and oil seeds, groundnut meal and soybean meal

OMd = 87.75 – 0.314 CF + α (n = 46; r = 0.74; RSD = 3.8)

Where α = -4.36 full fat soybeans

α = -1.86 groundnut meal

α = +6.2 other legume and oil seeds, soybean meal

##### Sugar beet pulp and citrus pulp

Sugar beet pulp:

OMd = 87.20 – 0.951 (CF – 16.39)

Citrus pulp:

OMd = 84.11 – 1.374 (CF – 16.39)

For the two groups: n = 34; r = 0.86; RSD = 2.0

##### Cassava, molasses, vinasse and potatoes

OMd = 97.81 – 1.12 NDF (n = 5; r = 0.98; RSD = 2.6)

##### Dehydrated forages

Alfalfa:

OMd = 65.90 – 0.919 (ADF – 29.83)

Grasses:

OMd = 74.13 – 1.364 (ADF – 29.83)

For the two groups: n = 32; r = 0.86; RSD = 3.6

In these equations, CF (crude fibre), NDF and ADF are expressed in % dry matter and OMd in %.

The residual standard deviations for these equations were between 2 and 4% of OMd. The equations were then applied to the chemical composition of the feed materials presented in the tables. With this method, we obtained consistent values for OMd for more than 85% of the feed materials. In the case of some rarely used by-products, no recent data of *in vivo* digestibility were available, and we derived their digestibility values from the tables of Becker and Nehring (1965), an old but well-documented source of feed information.

#### Calculation of fatty acid digestibility

In order to estimate the digestibility of fatty acids, we used an equation (Sauvant and Bas, 2001) based on a quantitative review of literature data:

FAduo = 0.83 FAfeed + 0.84 (n = 116; number of trials = 38; r = 0.97; RSD = 0.54)

FAduo is the duodenal flow of fatty acids (% dry matter intake) and FAfeed is the fatty acids content in the feed materials (% dry matter intake). The intestinal digestibility of fatty acids was assumed to be 75%.

Fatty acid digestibility (FAd in %) is calculated by:

FAd = 100 x (FAfeed – (1 – 0.75) x FAduo) / FAfeed

Or: FAd = 79 – (21 / (EE x FA/EE coefficient / 100))

EE is ether extract in % dry matter. The FA/EE coefficient is the percentage of fatty acids in the ether extract (expressed in %).

It should be noted that some digestibility values were negative, for example for low-fat feed materials, due to the fact that rumen microorganisms synthesize fats. These negative values were not published in the tables.

#### Calculation of energy values

The calculations of energy values were performed using the approach and the equations proposed by INRA (INRA, 1978; INRA, 1988; Vermorel *et al*., 1987). UFL and UFV values were calculated for each feed material in the tables: UFL (Unité Fourragère Lait, feed unit for milk production) is the energy unit for lactating ruminants or slow growing ruminants; UFV (Unité Fourragère Viande, feed unit for meat production) is the energy unit for medium to fast growing ruminants. One UF corresponds to the energy value of a kilogramme of standard barley.

##### Energy digestibility (Ed)

Ed was calculated from OMd and chemical components using regression equations obtained from the database cited above:

Ed = OMd – 3.94 + 0.104 CP + 0.149 EE + 0.022 NDF – 0.244 Ash

(n = 183; r = 0.68; RSD = 1.5)

Ed = OMd – 3.50 + 0.046 CP + 0.155 EE

(n = 216; r = 0.35; RSD = 1.8)

Ed = OMd – 2.90 + 0.051 CP

(n = 250; r = 0.35; RSD = 2.0)

Ed and OMd are expressed in %; the chemical constituents (CP: crude protein; EE: ether extract; Ash: minerals) are expressed in % dry matter.

The choice between the different equations is based on the availability of analytical data and the most precise equation should be used.

##### Digestible and metabolisable energy levels

Digestible energy: DE = GE x Ed / 100

Metabolisable energy: ME = DE x ME / DE

100 ME / DE = 86.38 – 0.099 CFo – 0.196 CPo

GE: gross enery in MJ/kg dry matter; CFo: crude fibre in % of organic matter; CPo: crude protein in % of organic matter.

Metabolisable energy concentration of the feed: q = ME / GE (0 < q < 1)

The coefficient q is used in the equations below.

##### Efficiency of utilisation of metabolisable energy for net energy

The coefficients kl, km and kf are comprised between 0 and 1.

For lactation: kl = 0.60 + 0.24 (q – 0.57)

For maintenance: km = 0.287 q + 0.554

For fattening: kf = 0.78 q + 0.006

For maintenance and meat production: kmf = __km x kf x 1.5__

kf + 0.5 x km

##### UFL and UFV values

UFL value = __ME x kl__

7.12

UFV value = __ME x kmf__

7.62

The NE values of the reference barley were kept at 7.12 and 7.62 MJ/kg for milk and meat production respectively, even though the most recent NE values for average barley are 6.76 and 7.08 MJ/kg as fed respectively.

#### Comparison to previous tables

The chemical composition and the digestibilities values in the present tables were compared to the ones in previous tables: the INRA 1988 tables, the Dutch tables from CVB (2000) and older German tables (Schiemann *et al*., 1971).

The proposed composition vectors are in general highly correlated with those that we published in 1988. However, there are some significant variations for certain feed materials. The organic matter digestibilities are well correlated with those from the previous French, Dutch and German tables. However, the regression residuals indicate that digestibility can vary significantly for some feed materials, with notable consequences on their energy values.

### Nitrogen values

The apparent nitrogen digestibility values (Nd) of the feed materials in the database were quite variable, as they had been calculated using the difference method. Therefore, the Nd values were calculated using the following prediction equation for undigested crude protein (UDCP in % dry matter):

UDCP = 0.519 + 0.178 CP + 0.095 UDOM – 0.036 ADF

(n = 575; r = 0.68; RSD = 0.84)

This equation was obtained using various diets for cattle in different physiological statuses. Crude protein (CP), undigested organic matter (UDOM) and ADF are in % dry matter (Sauvant and Mertens, unpublished).

The Nd values were then calculated by:

Nd = 100 x (CP – UDCP) / CP

The nutritional values for nitrogen are expressed as digestible protein in the intestine or PDI (Protéines digestibles dans l'intestin, in g/kg). Three PDI values are indicated (INRA, 1978 and INRA, 1988):

- PDIA, of feed origin, included in PDIE and PDIN.
- PDIE, when energy is the limiting factor for rumen microbial activity.
- PDIN, when nitrogen is the limiting factor for rumen microbial activity

#### Calculation of *in sacco* degradability

In order to update the data published in 1988, we collated the kinetics of nitrogen degradability measured in three laboratories: the URH at Theix (n = 112 samples and 32 feed materials), the UMR INRA-ENSAR Production du lait in Rennes (n = 76 samples and 21 feed materials) and the UMR INRA-INA P-G Physiologie de la nutrition et alimentation in Paris (n = 283 samples and 92 feed materials). Data from the Glon-Sanders laboratory (n = 113 samples and 29 feed materials), already used in the INRA 1988 tables, were used again for the diversity of these feed materials. For the same reason, literature data (139 references, 1305 samples and 178 feed materials) were also taken into account. The data were treated by analysis of variance to obtain, for each feed material, values for “effective degradability” of nitrogen (NED) corrected for laboratory bias. Similar approaches were used for the kinetic parameters a, b and c described below. For a certain number of ingredients that were absent or poorly represented in the databases, groups of reliable data were taken from the literature. This survey also provided original and useful information on the effects of some technological treatments.

The 3 basic parameters a, b and c for the kinetics of *in situ* nitrogen disappearance were included when the data was available:

nitrogen degradability (%) at time t: a + b x (1 - e^{-ct})

a: immediately degradable nitrogen fraction (%)

b: potentially degradable nitrogen fraction (%)

c: hourly rate of nitrogen particle degradation (h^{-1}).

The nitrogen effective degradability (NED in %) is calculated by:

NED = a + ((b x c)/(k + c))

Where k is the hourly rate of particle disappearance from the rumen, fixed at 0.06 h^{-1}.

#### Calculation of intestinal digestibility values for rumen undegradable protein

The values for true intestinal digestibility (TId) of rumen undegradable protein (RUP) were calculated when possible using the results obtained by the technique of mobile nylon bags in the intestine (UMR INRA-ENSAR). In order to increase the diversity of the feed materials, these data were completed by measurements performed using a similar methodology by around 15 other laboratories renowned for their reliability. To be included, each laboratory had to have tested at least 10 feed materials.

The 388 values collected (corresponding to 72 feed materials) were analysed by analysis of variance to obtain TId values corrected for laboratory bias. This also enabled the quantification of the effects of some technological treatments.

The quantities of RUP undigested in the intestine (RUPUI) were estimated by the nylon bag technique (RUPUI = RUP x (1 – Tid / 100)). We verified that this fraction corresponded with the true undigestible RUP by establishing a relationship between the quantities of crude protein not digested in the digestive tract [UDCP = CP x (1 – Nd / 100)] on one side, and the fermentable nitrogen fractions in the rumen [FCP = CP x NED / 100] not digested in the intestine, the digestible organic matter (DOM) and the indigestible organic matter (UDOM) on the other side:

UDCP = RUPUI + 0.24 FCP + 0.36 UDOM + 0.057 DOM

(n = 54; r = 0.87; RSD = 10.9)

UDCP, RUPUI, FCP, UDOM and DOM are expressed as % dry matter

The true digestibility in the intestine of RUP is calculated as TId = 100 x (RUP – RUPUI) / RUP.

For feed materials without a measured TId value, the digestibility was estimated from the equation:

TId = 88.3 + 0.371 CP – 0.0037 CP^{2} – 1.07 ADL – 0.313 UDOM

(n = 69; r = 0.95; RSD = 4.7)

TId is expressed in %.

CP and ADL are expressed in % dry matter.

The TId values obtained by this equation were rounded up or down to multiples of 5.

#### Calculation of PDI values for feed materials

The equations proposed in 1988 (Verité *et al*., 1987 and INRA, 1988) have been used to calculate the values of PDI for the different feed materials:

PDIA = CP x [1.11 (1 – NED)] x TId

PDIMN = CP x [1 – 1.11 (1 – NED)] x 0.9 x 0.8 x 0.8

PDIME = FOM x 0.145 x 0.8 x 0.8

PDIE = PDIA + PDIME

PDIN = PDIA + PDIMN

with CP: crude protein in g/kg dry matter

NED: nitrogen effective degradability in the rumen (0 < NED < 1)

TId: true digestibility of RUP in the small intestine (0 < TId < 1)

FOM: fermentable organic matter in the rumen, in g/kg dry matter, where FOM = digestible organic matter – ether extract – non degradable crude protein (*a priori* CP x (1 – NED))

When a feed material contained more than 10% starch, the level of FOM was reduced to take into account the starch fraction that escapes rumen digestion. The multiplication coefficient used for FOM was 0.6 for maize, sorghum and rice, and 0.8 for peas, faba beans, potato pulp, whole potatoes, full fat and extracted rice bran, maize germ meals, maize bran and cassava. A value of 0.95 was used for all the other feed materials containing more than 10% starch.

#### Digestible amino acids in the intestine

The values of digestible amino acids in the intestine were calculated using the method proposed by Rulquin *et al*. (2001a and b). The values indicated in the present tables are not exactly the same as the values published in previous tables since the values of crude protein, amino acids, nitrogen degradability and RUP digestibility (TId) have all been updated. Nine essential amino acids have been taken into account: lysine, methionine, leucine, histidine, phenylalanine, threonine, isoleucine, valine and arginine.

### Degradability of dry matter and starch

Kinetic parameters of *in sacco* dry matter degradation, which is often measured along with nitrogen degradation, have been indicated. The short-term degradability of dry matter is of particular interest as it gives an idea of the acidogenic activity of the feed material in the rumen. Finally, kinetic parameters of *in sacco* degradability of starch have been proposed for starch-rich feed materials, using the literature review by Offner *et al*. (2003), which updates the values previously published by Sauvant *et al.* (1994). Unlike the values for nitrogen and dry matter degradability, most of the starch degradability values were taken from a literature review and not from INRA studies. The starch degradation data can be used to estimate the flow of glucose absorbed in the intestine (Nocek and Tamminga, 1991; Sauvant *et al*., 1994; Sauvant, 1997), and also to modify the level of fermentable organic matter of feed materials containing large amounts of rumen undegraded starch, as has been done with the Dutch version of the PDI system (Tamminga *et al.,* 1994).

### Absorbed phosphorus

For the first time, values for absorbed phosphorus (g/kg) are published together with the classic total phosphorus values (g/kg) and INRA will propose a new system for dietary recommendations expressed in absorbed (net) phosphorus for ruminants. Only absorbed phosphorus can cover the net physiological requirements of animals. The data come from measurements of true phosphorus digestibility estimated by isotope dilution of ^{32}P (Compère, 1967; Grace, 1980; Field *et al.,* 1984; Challa and Braithwaite, 1989) or alternative methods (Aguerre *et al.,* 2002; Bravo *et al.,* 2002), and from indirect calculations using a database of literature results established from around 100 publications. Within a family of feed materials (cereals, oil seed meals), the specific results from each feed material were used when available. When this was not possible, the final value was an average value weighted by the number of results available for each feed material in the family concerned. In all cases we applied a safety margin taking into account the variability of the measurements. These data will soon be completed by digestible phosphorus values for forages in order to make the system fully operational.

### References

- Aguerre M. J., Marçot S., Henselmeyer H., Satter L. D., 2002. Availability of phosphorus in dairy feeds.
*J. Dairy Sci. supp.*(in press) - Becker M., Nehring K., 1965.
*Handbuch der Futtermittel*. Ed. Verlag Paul Parey , Berlin. - Bravo D., Bogaert C., Meschy F., Sauvant D., 2002. Availability of phosphorus from several meals for ruminants.
*Anim. Feed Sci.**Technol.*(in press) - Challa J., Braithwaite G. D., 1989. Phosphorus and calcium metabolism in growing calves with special emphasis on phosphorus homeostasis. 4. Studies on mil-fed calves given different amounts of dietary phosphorus with a constant intake of calcium.
*J. Agric.**Sc. (Camb.).,*113, 285-289. - Compère R., 1967. Influence de la nature de l’aliment sur le métabolisme phosphoré. In : Duculot, J. (Ed)
*L’étude de l’équilibre phosphocalcique du régime basé sur les variations du métabolisme phosphoré.*Les presses agronomiques de Gembloux. 109-113. - CVB, 2000. Veevoedertabel 2000.
*Chemische samenstelling, verteerbaarheid en voederwaarde van voedermiddelen*, Centraal Veevoederbureau, Lelystad, Pays-Bas. - Field A. C., Wooliams J. A., Dingwall R. A., Munro C. S., 1984. Animal and dietary variation in the absorption and metabolism of phosphorus by sheep.
*J. Agric.**Sci.,*103, 283-291. - Grace N. D. 1980. Effect of increasing phosphorus intake on the phosphorus faecal endogenous loss in sheep. Proc.
*N.Z. Soc.**Anim. Prod.,*40, 221-228. - INRA, 1978.
*Alimentation des ruminants.*INRA Editions, Versailles. - INRA, 1988.
*Alimentation des bovins, ovins et caprins.*INRA Editions, Paris, 471 p. - Nocek J. E., Tamminga S., 1991. Site of digestion of starch in the gastrointestinal tract of dairy cow and its effect on milk yield and composition.
*J. Dairy. Sci.*, 74, 3598-3629. - Offner A., Bach A., Sauvant D., 2003. Quantitative review of in situ starch degradation in the rumen.
*Anim. Feed Sci.**Technol.,*106, 81-93. - Rulquin H., Vérité R., Guinard-Flament J., 2001. Acides aminés digestibles dans l’intestin. Le système AADI et les recommandations d’apport pour la vache laitière.
*Prod. Anim.,*14, 265-274. - Rulquin H., Vérité R., Guinard-Flament J., Pisulewski P. M., 2001. Acides aminés digestibles dans l’intestin. Origine des variations chez les ruminants et répercussion sur les protéines du lait,
*Prod. Anim.*, 14, 201-210. - Sauvant D., 1997, Conséquences digestives et zootechniques des variations de la vitesse de digestion de l'amidon chez les ruminants.
*Prod. Anim.*, 10, 287-300. - Sauvant D., Bas P., 2001. La digestion des lipides chez le ruminant,
*Prod. Anim.*, 14, 303-310. - Sauvant D., Chapoutot P., Archimède H., 1994. La digestion des amidons par les ruminants et ses conséquences.
*Prod. Anim.*, 7, 115-124. - Schiemann R., Nehring K., Hoffman L., Jentsch W., Chudy A., 1971.
*Energetische Futterbewertung und Energienormen*, VEB Deutscher Landwirtschaftsverlag, Berlin, 344 p. - Tamminga S., Van Straalen W. M., Subnel A. P. J., Meijer R. G. M., Steg A., Wever C. J. G., Block M. C., 1994. The Dutch protein evaluation system: the DVE/OEB system.
*Livest. Prod. Sci.*, 40, 139-145. - Vérité R., Michalet-Doreau B., Chapoutot P., Peyraud J. L., Poncet C., 1987. Révision du système des protéines digestibles dans l’intestin (PDI),
*Bull.**Tech. CRZV*, Theix, INRA. 70, 19-34. - Vermorel M., Coulon J. B., Journet M. R., 1987. Révision du système des Unités Fourragères (UF),
*Bull.**Tech. CRZV*, INRA. 70, 9-18.