Variation between individual cows in in situ rumen degradation characteristics of maize and grass silages

Highlights

⿢	Between-cow variation is essential to obtain accurate estimate of rumen degradation.
⿢	Pooling of rumen incubated residues depends on variation between individual cows.
⿢	Three cows are sufficient for in situ rumen incubations of grass silages.
⿢	Four or more cows are required for in situ rumen incubations of maize silages.

Abstract

Different numbers of animals have been used in different studies to cover the variation between individual animals in in situ rumen degradation characteristics of maize and grass silages. The objective of this study was to determine whether three cows are sufficient or not to cover the variation between individual cows in in situ rumen degradation characteristics of dry matter (DM), organic matter (OM), crude protein (CP), starch and neutral detergent fibre (NDF) for maize and grass silages. Fifteen maize and 15 grass silage samples, with a broad range in chemical composition, were selected. The maize and grass silage samples were incubated in the rumen for 2, 4, 8, 16, 32, 72 and 336 h, using the nylon bag technique. Three cows were used for nylon bag incubation of maize silages and three other cows for grass silages. The variation between individual cows was found significant (P < 0.05) for degradation rate (k_d) of DM, OM and CP, and the effective rumen degradation (ED) of DM and CP of maize silages whereas non-significant (P > 0.05) differences were found for all other parameters of DM, OM, CP, starch and NDF. The variation between individual cows was found non-significant (P > 0.05) for rumen undegradable fraction (U), potentially rumen degradable fraction (D), k_d and ED of DM, OM, CP and NDF of grass silages. The results of this study indicate that the use of three cows for nylon bag incubations of grass silages is sufficient whereas the use of three cows for nylon bag incubations of maize silages is not sufficient to cover variation between the individual cows.

Keywords:

• Maize silage
• Grass silage
• Variation between cows
• Effective rumen degradability
• Rumen degradation characteristics

1 Introduction

Dietary nutrient bioavailability of feed or feed ingredient is essential information to ensure that the nutrient requirements of animals are met. Several in situ, in vitro and in vivo techniques have been developed to measure nutrient digestibility and over the past decades, much research has focussed on determining the nutritive value of major forages and concentrate ingredients used in dairy cow nutrition. The in situ (in sacco) technique has been extensively used for the determination of the rumen degradation of different chemical components in the rumen [1⿿3]. Factors affecting the results of in situ technique include bag and pore size, sample size, bag material, bag insertion and removal procedures, rumen incubation time, number of replicate animals, animals, diet of the experimental animals, feeding level, feeding frequency, rinsing procedure, mathematical models, and microbial contamination [4,5]. A number of these factors have been extensively studied [6⿿8]. Less focus has been directed on investigating the variation between individual animals for rumen incubations of forages.

In the past, many studies have been conducted to determine the in situ rumen degradability of dietary fractions or nutrients of maize and grass silages. In many of these studies, an arbitrary assumption was made that two [9,10], three [11⿿13], four [14,15] or six [16,17] animals are sufficient for in situ rumen incubations of forages. However, little data exists in the scientific literature to determine the validity of this assumption for in situ rumen incubations of forages. Various ruminant feed evaluation systems such as the DVE/OEB₂₀₁₀ system [18] in the Netherlands and the NorFor ⿿ the Nordic feed evaluation system [19] recommend to use three cows for rumen incubations of feeds and allow pooling of rumen incubated residues to perform chemical analysis. The use of three cows for in situ rumen incubations has not been validated for maize and grass silages. In the past, few studies were performed with other feed ingredients to determine the variation between individual cows. Weakley et al. [6] observed no significant differences between individual animals (n = 4), days of incubation and periods of experimentation on ruminal disappearance of dry matter (DM) of soybean meal whereas significant differences between individual animals in in situ nitrogen (N) disappearance were observed. Figroid et al. [20] found significant differences between steers (n = 2) in in situ rumen DM degradability for barley and sorghum. Recently, an in situ study conducted by Castillo-Gallegos et al. [8] with king grass (Pennisetum purpureum) leaves showed no significant differences between cows (n = 3) for DM disappearance and concluded that two cows are sufficient for in situ rumen incubations. As the experimental costs increase significantly with increasing animal numbers, therefore, knowledge of between-animal variation is essential to obtain cost-effective estimates for in situ rumen degradation of maize and grass silages.

This study was designed to determine whether three cows are sufficient or not to cover the variation between the individual cows in in situ rumen degradation characteristics of DM, organic matter (OM), crude protein (CP), neutral detergent fibre (NDF) and/or starch of maize and grass silages.

2 Materials and methods

2.1 Samples collection and processing

Fifteen maize and 15 grass (mainly Lolium perenne) silage samples (⿼5 kg per silage) were obtained during 2007, 2008 and 2009 from various commercial farms located in different regions in the Netherlands. The samples were collected by trained technicians from a feed analysis laboratory (Blgg Research, Wageningen, The Netherlands) using a hollow drill. After collection, the samples were stored at ⿿20 °C until processing. The frozen samples were cut using a bread slicer (JAC Duro BEL 450; ABO, Leek, The Netherlands) having a distance of 11 mm between the discs, thoroughly mixed by hand and divided into three parts; one part (⿼2.5 kg) was subjected to chemical analyses after freeze drying and grinding (3 mm), another part (⿼1.5 kg) was stored at ⿿20 °C for later in situ rumen incubations, and the third part (⿼1.0 kg) was stored (⿿20 °C) as a reserve for possible future analysis.

2.2 In situ rumen incubations

Six multiparous (second or third lactation) Holstein Friesian cows, producing >15 kg milk per day and fitted with permanent rumen cannulas, were used in this experiment. Cows were fed a total mixed ration (Table 1) and had 24 h/day access to fresh water. Information on age, lactation number, milk production, milk protein, and milk fat production of the six cows used for the maize and grass silage rumen incubations is presented in Table 2. Three cows (Cow 1⿿3) were used for in situ rumen incubations of maize silages whereas other three cows (Cow 4⿿6) were used for grass silages. The maize and grass silage samples (⿼5 g DM) were weighed into 10 cm ÿ 19 cm nylon bags (porosity 24%; pore size 37 μm; Nybolt, Zürich, Switzerland) and incubated in the rumen for 2, 4, 8, 16, 32, 72 and 336 h according to the procedure used by Ali et al. [21]. The 0 h bags were washed in a washing machine (AEG-Electrolux ÿko Turnamat 2800, Stockholm, Sweden) for 40 min using tap water at 25 °C without incubation in the rumen [22]. The washed bags were stored at ⿿20 °C for at least 24 h, subsequently freeze dried and the residues were used to calculate the washout (W) fraction. Six bags of each maize and grass silage sample were incubated in the rumen of the three cows (2 bags per cow per incubation time) for 2, 4, 8, 16, and 32 h. Because of the low recovery of incubated residues per nylon bag for the 72 and 336 h incubation periods, 9 bags of each silage sample were incubated in the rumen of the three cows (3 bags per cow per incubation time) for these incubation periods. After removal from the rumen, bags were frozen at ⿿20 °C for at least 24 h, after which the bags were thawed and washed in the washing machine as described above. The washed bags were stored at ⿿20 °C and subsequently freeze dried. Rumen incubation residues of each sample collected from each cow after each rumen incubation period were pooled and ground separately over a 1 mm sieve, using a hammer mill (Pepping, 200 AN-797002, Deventer, The Netherlands).

Table 1 Components of the total mixed ration fed to the cows during rumen incubations.

Feed ingredient	g/kg DM
Maize silage	237.4
Grass silage	395.8
Grass hay/wheat straw	13.5
Soybean meal	47.4
Sweet syrup	9.5
Wet distillers grains with solubles	84.3
Soybean meal (rumen protected)	18.6
Vitamin and minerals premix^a	10.9
Concentrate^b	182.6

a Contained: calcium 175 g/kg; phosphorous 0.2 g/kg; magnesium 130 g/kg; sodium 50 g/kg; chloride 78 g/kg; vitamin A 600,000 IE/kg; vitamin D 120,000 IE/kg; vitamin E 8000 IE/kg.

b Contained: maize, 30%; palm kernel, 22%; rapeseed solvent extract, 20%; citrus pulp, 10%; soybean meal (rumen protected), 5%; Beet molasses, 5%; molasses, 2%; wheat, 2%; chalk (CaO), 0.6%; urea, 0.6%; salt, 0.5%; Magnesium oxide (80% MgO, 0.5%; palm oil, 0.2%; vitamin and minerals premix, 0.3%.

Table 2 Information about the cows used for maize and grass silage rumen incubations.

Variable	Maize silage rumen incubations			Grass silage rumen incubation
	Cow 1	Cow 2	Cow 3	Cow 4	Cow 5	Cow 6
Age (days)	1754	1717	1654	1388	1397	1271
Lactation number	2	3	2	2	2	2
Milk production/day (kg)	29.7	27.2	30.9	28.0	29.7	27.6
Milk protein/day (kg)	1.1	1.1	1.0	1.0	1.1	1.0
Milk fat/day (kg)	1.5	1.4	1.2	1.2	1.3	1.3

2.3 Chemical analysis and calculations

All the ground freeze dried samples were analysed for DM, ash, CP, crude fat, sugar, NDF, acid detergent fibre and acid detergent lignin. Incubation residues were analysed for DM, ash, CP and NDF. Additionally, starch was analysed in maize silage samples and rumen incubated residues of maize silages.

The DM content was determined by oven drying at 103 °C for 4 h (ISO 6496) and ash content by incineration at 550 °C for 4 h (ISO 5984). The N content was determined using the Kjeldahl method (ISO 5983) and CP was calculated as N ÿ 6.25. Starch was determined using the amyloglucosidase method (ISO 15914) after dissolving in 100% dimethyl sulfoxide. The NDF was determined according to the modified method of Van Soest et al. [23], using amylase and expressing without residual ash (ISO 16472). Acid detergent fibre was analysed by boiling with acid detergent reagent and expressed without residual ash (ISO 13906:2008). Acid detergent lignin was determined after boiling with acid detergent reagent and solubilisation of cellulose with sulphuric acid (ISO 13906:2008). The crude fat was determined by ISO 6492 and sugar content was determined by the Luff-Schoorl method (NEN 3571:1947nl).

The fractional degradation rate (k_d) of DM (k_d-DM), OM (k_d-OM), CP (k_d-CP) and starch (k_d-starch) was calculated according to the first order model of Robinson et al. [24] including U, D and k_d:

(1) Y_t = U + D_t ÿ exp (⿿k_d ÿ t)

where Y_t = degradation at time t, U = the rumen undegradable fraction, D_t = fraction potentially degradable in the rumen at time t, t = time of incubation (h).

The k_d of NDF (k_d-NDF) was calculated by the same model including a lag time:

(2) Y_t = U + D_t ÿ exp (- k_d ÿ (t ⿿ L))

where L = lag time.

The effective rumen degradability (ED) of DM (ED_DM), OM (ED_OM), starch (ED_starch) and NDF (ED_NDF) was calculated according to the equation of ÿrskov and McDonald [25]:

ED = W + (k_d/(k_d + k_p)) ÿ D

where W = washout fraction, k_p = fractional passage rate (h^⿿1) and D = potentially rumen degradable insoluble fraction. The ED_DM and ED_OM of maize and grass silages were calculated using an assumed k_p value of 0.05 h^⿿1. For maize silages, a k_p value of 0.06 h^⿿1 was used for starch, based on Van Duinkerken et al. [19]. For maize and grass silages, k_p values of 0.020 h^⿿1 and 0.025 h^⿿1, respectively, were used for NDF, based on data of Pellikaan [26]. For forages, 5% of the W fraction of CP was assumed to escape rumen degradation [18] and is considered to be part of the rumen escape protein (REP). Therefore, the ED of CP (ED_CP) and REP were calculated by the modified equations;

ED_CP = 0.95 ÿ W + (k_d/(k_d + k_p)) ÿ D

REP = U + (k_p/(k_d + k_p)) ÿ D + 0.05 ÿ W

The k_p value of 0.045 h^⿿1 for CP was adopted from the DVE/OEB₂₀₁₀ system [18].

2.4 Statistical analysis

Rumen degradable data of DM, OM, CP, starch and NDF of maize and grass silages after incubation in the rumen of three cows for 2, 4, 8, 16, 32, 72 and 336 h were summarized by descriptive statistics. The k_d-DM, k_d-OM, k_d-CP and k_d-starch of maize silages and the k_d-DM, k_d-OM, and k_d-CP of grass silages were calculated using model 1 and the k_d-NDF of maize and grass silages was calculated using model 2 in SAS^® [27]. Variation between individual cows was determined for U, D, k_d and ED of DM, OM, CP, starch and NDF of maize silages, and the DM, OM, CP and NDF of grass silages by PROC MIXED procedure of SAS^® [28]. Variation between individual cows for different incubation periods was determined by PROC MIXED procedure of SAS^® [27].

Data were analysed for the fixed effect of cow and random effect of samples using the PROC MIXED procedure of the SAS.

Y_ijk = μ + PS_i+S_j + п_ijk

where, Y_ijk is the dependent variable; μ the general mean; PS_i is the fixed effect of cows i (i = 1⿿3); S_j is the random effect of samples j (j = 1⿿15 for maize silage samples and 1⿿15 for grass silage samples); п_ijk is the residual. Values were considered to differ significantly when P < 0.05.

3 Results

The chemical composition of the maize and grass silage samples is presented in Table 3. A broad range was found in the chemical components of maize and grass silages. The DM content of maize silages ranged from 272.2 to 478.8 g/kg DM whereas in grass silages it ranged from 270 to 685 g/kg DM. Higher range was found in CP content (102⿿220 g/kg DM) of grass silages compared to CP (55.1⿿81 g/kg DM) of maize silages

Table 3 Chemical composition and silage quality parameters of maize and grass silages.

Variable	Maize silage (n = 15)				Grass silage (n = 15)
	Mean	S.D^a	Minimum	Maximum	Mean	S.D	Minimum	Maximum
Chemical composition (g/kg DM)^d
Dry matter (g/kg fresh matter)	348.1	49.3	272.2	478.8	464.9	126.7	237.0	685.0
Ash	34.0	6.4	22.0	43.0	109.6	19.8	85.5	163.0
Crude protein	68.2	8.1	55.1	81.0	158.9	32.9	102.0	220.0
Crude fat	36.5	5.1	29.0	47.0	40.5	9.5	27.0	65.0
Crude fibre	186.9	24.3	138.0	228.0	281.7	22.0	243.0	316.0
Starch	339.8	36.7	286.0	402.0	nd^b	nd	nd	nd
Sugar	10.0	4.1	5.8	22.0	62.6	37.6	11.0	135.0
Neutral detergent fibre	394.7	37.9	326.0	451.0	545.4	44.5	464.0	611.0
Acid detergent fibre	218.2	25.4	180.0	263.0	297.1	20.1	254.0	316.0
Acid detergent lignin	19.4	3.6	14.0	26.0	23.2	7.1	14.0	39.0
Silage quality parameters^e
pH	3.89	0.17	3.60	4.30	5.14	0.63	4.23	6.12
NH3-nitrogen (g N/kg DM)^c	5.72	1.86	1.58	8.66	2.62	1.28	0.86	5.68

a Standard deviation.

b Not determined.

c NH₃-nitrogen determined on fresh material and then recalculated to content in DM of air dried material.

d Determined on freeze dried material.

e Determined on air dried material.

3.1 Maize silages

The average values for U, D, k_d and ED of DM, OM, CP, starch and NDF of the 15 maize silages for individual cows are presented in Table 4. Standard error of the mean (SEM) for k_d-CP was found higher compared to the k_d values of DM, OM, starch and NDF of the maize silages (Table 4). The range in the average values of ED_DM was larger compared to those for ED_OM, ED_CP, ED_starch and ED_NDF of the maize silages. The range in average values for REP of maize silages between the three cows was 0.418⿿0.436 (Table 4). The lag time for NDF degradation of maize silages for the three cows ranged from 4.88 to 5.55 h (Table 4). The W fraction for DM, OM, CP, starch and NDF of maize silages was 0.281, 0.281, 0.519, 0.445 and 0.002, respectively. There was a significant difference between individual cows for k_d-DM (P = 0.02), ED_DM (P = 0.01), k_d-OM (P = 0.03), k_d-CP (P = 0.005), ED_CP (P < 0.001) and REP (P < 0.001) of the maize silages whereas non-significant difference (P > 0.05) was observed for all others parameters of DM, OM, CP, starch and NDF. The average rumen degradability of DM, OM, CP, starch and NDF after rumen incubation for 2, 4, 8, 16, 32, 72, and 336 h are presented in Fig. 1. Differences in rumen degradability between individual cows were found significant for 16 and 72 h rumen incubation periods (P < 0.01).

Fig. 1 Average rumen degradability of (A) dry matter, (B) organic matter, (C) crude protein, (D) starch and (E) neutral detergent fibre of 15 maize silages after incubated for 2, 4, 8, 16, 32, 72, and 336 h in the rumen of cow 1 (

), 2 (

) and 3 (

). Error bar represents ± standard error of the mean. **, (0.001 < P < 0.01) and ***, (P<0.001) indicate significant differences between the cows.

Table 4 Average values for rumen undegradable (U) fraction, potentially rumen degradable (D) fraction, degradation rate (k_d, h^⿿1) and effective rumen degradability (ED)^a of maize silage (n = 15) for three cows.

Variable	Cow 1	Cow 2	Cow 3	Mean	SEM^b	P value	Significance^c
Dry matter (DM)
UDM	0.162	0.165	0.173	0.167	0.003	0.31	NS
DDM	0.557	0.554	0.546	0.552	0.003	0.31	NS
kd-DM	0.042	0.039	0.046	0.042	0.002	0.02	*
EDDM	0.531	0.521	0.542	0.531	0.006	0.01	*
Organic matter (OM)
UOM	0.164	0.176	0.179	0.173	0.005	0.25	NS
DOM	0.556	0.544	0.540	0.547	0.005	0.25	NS
kd-OM	0.042	0.039	0.046	0.042	0.002	0.02	*
EDOM	0.532	0.518	0.537	0.529	0.006	0.08	NS
Crude protein (CP)
UCP	0.297	0.286	0.299	0.294	0.004	0.15	NS
DCP	0.184	0.196	0.182	0.187	0.004	0.15	NS
kd-CP	0.029	0.021	0.035	0.028	0.004	0.005	**
EDCP	0.575	0.564	0.582	0.574	0.005	0.008	***
REP^d	0.425	0.436	0.418	0.426	0.005	0.008	***
Starch
Ustarch	0.005	0.004	0.003	0.004	0.001	0.27	NS
Dstarch	0.550	0.551	0.552	0.551	0.001	0.27	NS
kd-starch	0.101	0.096	0.096	0.098	0.002	0.63	NS
EDstarch	0.786	0.779	0.773	0.779	0.004	0.62	NS
Neutral detergent fibre (NDF)
UNDF	0.247	0.258	0.263	0.256	0.005	0.17	NS
DNDF	0.751	0.740	0.736	0.742	0.004	0.17	NS
kd-NDF	0.026	0.026	0.029	0.027	0.001	0.50	NS
Lag time (h)	4.882	5.553	4.883	5.106	0.224	0.82	NS
EDNDF	0.413	0.413	0.423	0.416	0.003	0.60	NS

a The ED_DM and ED_OM was calculated using a value of 0.05 h^⿿1 for the passage rate. The ED_CP, ED_starch and ED_NDF were calculated using passage rate values of 0.045 h^⿿1, 0.06 h^⿿1 and 0.020 h^⿿1, respectively.

b Standard error of mean.

c NS, not significant (P > 0.05); *, P < 0.05; **, P < 0.01; ***, P < 0.001.

d Rumen escape protein. 5% of the W fraction was added in REP.

3.2 Grass silages

The average values for individual cows for U, D, k_d and ED of DM, OM, CP and NDF of 15 grass silages are shown in Table 5. Variation between individual cows was found non-significant (P > 0.05) for U, D, k_d and ED of DM, OM, CP and NDF of grass silages. The between-cow variation in the k_d and ED values was somewhat larger for NDF compared to those of DM, OM and CP of grass silages. The range in average REP values of 15 grass silage samples for three cows was 0.433⿿0.435 (Table 5). The average lag time for NDF degradation was 5.71, 4.42 and 4.18 h for cow 4⿿6, respectively (Table 5). The W fraction for DM, OM, CP and NDF of grass silages was 0.193, 0.149, 0.383 and 0.008, respectively. The average in situ rumen degraded DM, OM, CP, and NDF of the 15 grass silages after incubation in the rumen of three cows for 2, 4, 8, 16, 32, 72, and 336 h are presented in Fig. 2. The differences in rumen degradability between individual cows were found significant for CP at 4 h (P = 0.006) and NDF at 2 h (P = 0.04). The differences between the cows in DM rumen degradability of grass silages were significant at 2 h (P = 0.052).

Fig. 2 Average rumen degradability of (A) dry matter, (B) organic matter, (C) crude protein and (D) neutral detergent fibre of 15 grass silages after incubated for 2, 4, 8, 16, 32, 72, and 336 h in the rumen of cow 4 (

), 5 (

) and 6 (

). Error bar represents ± standard error of the mean. *, (0.01 < P < 0.05) and **, (0.001 < P < 0.01) indicate significant differences between the cows.

Table 5 Average values for rumen undegradable (U) fraction, potentially rumen degradable (D) fraction, degradation rate (k_d, h^⿿1) and effective rumen degradability (ED)^a of grass silage (n = 15) for three cows.

Variable	Cow 4	Cow 5	Cow 6	Mean	SEM^b	P value	Significance^c
Dry matter (DM)
UDM	0.208	0.213	0.215	0.212	0.002	0.44	NS
DDM	0.599	0.594	0.592	0.595	0.002	0.44	NS
kd	0.040	0.040	0.041	0.040	0.000	0.97	NS
EDDM	0.448	0.453	0.449	0.450	0.002	0.90	NS
Organic matter (OM)
UOM	0.203	0.210	0.210	0.208	0.002	0.48	NS
DOM	0.648	0.641	0.641	0.643	0.002	0.48	NS
kd	0.037	0.039	0.040	0.039	0.001	0.87	NS
EDOM	0.416	0.428	0.423	0.422	0.003	0.63	NS
Crude protein (CP)
UCP	0.297	0.292	0.285	0.291	0.003	0.38	NS
DCP	0.320	0.325	0.332	0.326	0.003	0.38	NS
kd	0.073	0.074	0.076	0.074	0.001	0.96	NS
EDCP	0.565	0.566	0.567	0.566	0.001	0.95	NS
REP^d	0.435	0.434	0.433	0.434	0.001	0.95	NS
Neutral detergent fibre (NDF)
UNDF	0.211	0.218	0.216	0.215	0.002	0.82	NS
DNDF	0.781	0.774	0.776	0.777	0.002	0.82	NS
kd	0.039	0.052	0.036	0.042	0.005	0.41	NS
Lag time (h)	5.707	4.420	4.180	4.769	0.474	0.08	NS
EDNDF	0.480	0.482	0.470	0.477	0.004	0.24	NS

a The ED_DM and ED_OM was calculated using a value of 0.05 h^⿿1 for the passage rate. The ED_CP and ED_NDF were calculated using passage rate values of 0.045 h^⿿1 and 0.025 h^⿿1, respectively.

b Standard error of mean.

c Not significant (P > 0.05).

d Rumen escape protein. 5% of the W fraction was added in REP.

4 Discussion

Factors affecting the in situ rumen degradation characteristics of DM, OM, CP, starch and NDF of different feedstuffs have been extensively studied. There are much less scientific publications, however, reporting the variation between individual cows in rumen degradation characteristics of maize and grass silages. Authoritive feeding tables such as those of the CVB in the Netherlands and the NorFor⿿the Nordic feed evaluation system suggest to use three animals for in situ rumen incubations and pooling of rumen incubated residues before chemical analysis [19,28]. Information on the variation between animals is important for providing reliable estimates which can be used with confidence in practical feed formulation. Therefore, the present study was conducted to check whether three cows are sufficient for maize and grass silage rumen incubations.

4.1 Maize silages

The difference between individual cows in the present study was significant (P < 0.05, P < 0.01 and P < 0.001) for a number of parameters (k_d-DM, ED_DM, k_d-OM, k_d-CP and ED_CP) in the case of maize silages. Others researchers have also found significant differences between the cows in rumen degradation of dietary nutrients [6,20]. The average values for D_DM, k_d-DM and ED_DM of maize silages in the present study were 0.557, 0.042 h^⿿1 and 0.531, respectively. Ranjbari et al. [9] reported values of 0.489, 0.040 h^⿿1 and 0.480 for D_DM, k_d-DM and ED_DM for maize silages, respectively. The differences in the D_DM and ED_DM between the two studies may be due to the high average DM content (348 g/kg fresh matter) of the maize silages in the present study compared to those used (234 g/kg fresh matter) by Ranjbari et al. [9]. In the present study, the rumen incubated residues of CP were not corrected for microbial contamination which can affect the in situ results. However, microbial contamination is one of the main factors that affect the results of the in situ CP degradation of forages [29,30].

Differences between individual cows in rumen degradability were also assessed for specific rumen incubation periods. There were significant differences in rumen degradability between individual cows for DM, OM, CP and NDF of maize silages at rumen incubation periods of 16 and 72 h. The starch rumen degradability of maize silages between individual cows was non-significant (P > 0.05) at all rumen incubation time periods. The significant differences between the cows in CP rumen degradation for individual samples of maize silages indicate that more than three cows are required for in situ rumen incubations of maize silages. Moreover, the results of the present study do not suggest pooling of the rumen incubated residues of maize silages from three cows for performing chemical analysis.

4.2 Grass silages

For grass silages, non-significant (P > 0.05) differences between cows were found for U, D, k_d and ED of DM, OM, CP and NDF. Weakley et al. [6] also found non-signficant differences (P > 0.1) between in situ DM disappearance between the four cows whereas significant differences (P < 0.1) in in situ N disappearance for other feed ingredient (soybean meal). Castillo-Gallegos et al. [8] reported a smaller variance between cows for in situ DM disappearance of king grass leaves. Similar results for between cow variation for grass silages of the present study and the kinggrass leaves used by the Castillo-Gallegos et al. [8] indicates that three cows can be used for rumen incubations of grass and grass silages.

In the present study, the average values of the three cows for D_CP, k_d-CP and ED_CP for grass silages were 0.320, 0.073 h^⿿1 and 0.565, respectively. Von Keyserlingk et al. [17] reported values of 0.207, 0.088 h^⿿1 and 0.330 for D_CP, k_d-CP and ED_CP, respectively. The differences in the values between both studies could be due to the different average CP content of the grass silages (159 vs. 131 g/kg DM) as a high CP content of forages results in higher ED_CP values [31]. The rumen degradability of CP of grass silages also shows significant (0.01 < P < 0.05) differences between individual cows at 4 h, which might be due to the microbial contamination of CP residues. The differences between the cows in rumen degradability of DM, OM, CP and NDF of grass silages were small at specific rumen incubation periods.

5 Conclusions

The results of the present study suggest that three cows are sufficient for rumen incubations of grass silages under the present in situ conditions. The residues of three cows after each rumen incubation period can be pooled before chemical analysis. For maize silages, a number of estimates related to DM, OM and CP were significantly different between individual cows under the present in situ conditions. A study with four, five or six cows is recommended to find the exact number of cows required for rumen incubations of maize silages to cover the variation between cows.

Acknowledgements

The Higher Education Commission (HEC, Pakistan), the Dutch Product Board Animal Feed (PDV, Zoetermeer, The Netherlands) and the Dutch Dairy Board (PZ, Zoetermeer, The Netherlands) for financial support and BLGG Research (Wageningen, The Netherlands) for performing the chemical analysis.