| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ANIMAL NUTRITION |
The Agricultural Research Institute of Northern Ireland, Hillsborough, BT26 6DR Northern Ireland
Abstract
Grass silages (n = 136) were selected from commercial farms across Northern Ireland according to their pH, ammonia nitrogen, DM, and predicted ME concentration. Each silage was offered to four sheep as a sole feed at maintenance feeding level to determine nutrient digestibility and urinary energy output. Dry matter concentration was determined as alcohol-corrected toluene DM and was subsequently used as the basis for all nutrient concentrations. The objectives were to use these data to examine relationships between nutritive value and nutrient concentration or fermentation characteristics in silages and then develop prediction equations for silage nutritive values using stepwise multiple regression techniques. The silages had a large range in quality (DM = 15.5 to 41.3%, ME = 7.7 to 12.9 MJ/kg of DM, pH = 3.5 to 5.5) and a relatively even distribution over the range. There was a positive relationship (P < 0.001) between silage GE and DE or ME concentration. Digestible OM in total DM (DOMD); ME/GE; and digestibility of DM, OM, and GE were positively related (P < 0.05) to CP, soluble CP, ether extract, lactic acid concentration, and lactic acid/total VFA, whereas they were negatively related (P < 0.05) to ADF, NDF, lignin, individual VFA concentration, pH, and ammonia N/total N. Concentrations of DE and ME and digestibility of CP and NDF had similar relationships with those variables, although some relationships were not significant. Three sets of multiple prediction equations for DE and ME concentration; ME/GE; DOMD; and digestibility of DM, OM, GE, CP, and NDF were therefore developed using three sets of predictors. The first set included GE, CP, soluble N/total N, DM, ash, NDF, lignin, lactic acid/total VFA, and ammonia N/total N; the second set excluded soluble N/total N and lignin because they are not typically measured; the third set further excluded the fermentation data. The R2 values generally decreased with exclusion of predictors. The second and third sets of equations, except for NDF digestibility, were validated using the mean-square-prediction-error model and an independent grass silage data set published since 1977 (n = 17 [DM digestibility] to 28 [DOMD and OM digestibility]). The validation indicated that the equations developed in the present experiment could accurately predict DE and ME concentrations and DE/GE and ME/GE in grass silages.
Key Words: Digestibility • Energy Concentration • Grass Silage • Prediction Equation • Validation
Introduction
Dietary ME concentration is the basal unit in energy feeding systems currently adopted across the world (Agnew and Yan, 2000
). It is measured in digestibility trials (for energy digestibility and urinary energy output) and in calorimetric chambers (for methane energy). Alternatively, as recommended in many energy feeding systems, it can be predicted from digestibility of GE, OM, CP, and NDF and digestible OM in total DM (DOMD; Agnew and Yan, 2000). However, both digestibility trials and calorimetric measurements are labor-intensive and expensive approaches. It is thus unrealistic in practice to measure the ME concentrations for all concentrate supplements and forages used. The problem is especially difficult for grass silages as nutritive value in grass silages varies greatly according to grass variety, harvesting date, grass DM concentration, and ensiling technique. However, the nutritive value of grass silage is related to its nutrient concentration and fermentation characteristics. For example, DOMD in clamp grass silages was reported to be negatively related to the fiber fractions and ash concentration whereas positively related to CP concentration (Givens et al., 1989; Steg et al., 1990; Nousiainen et al., 2003). Similar negative relationships between DOMD and fiber fractions were also observed in big bale grass silages (Givens et al., 1993a). Therefore, it might be possible to use silage chemical analysis data to develop prediction equations for nutrient digestibility and ME concentration that can be used in practice. The objectives of the present experiment were to use grass silage data obtained at the Agricultural Research Institute of Northern Ireland to develop prediction equations for nutrient digestibility and energy concentration from nutrient concentrations and fermentation characteristics, and then validate the accuracy of these equations using grass silage data published since 1977.
Materials and Methods
Silages
A total of 136 grass silages produced on 125 commercial farms across Northern Ireland were used in the present experiment. These silages were selected according to their pH, ammonia N as a proportion of total N, DM, and predicted ME concentrations. The predicted ME concentration was estimated from the DOMD in silage, which was predicted by near-infrared reflectance spectroscopy (Barber et al., 1990). The objective in selecting silages was to obtain a large range in quality, with a relatively even distribution over the whole range. The silages were made from perennial ryegrass dominant swards and encompassed primary growth and first and second regrowth materials. The grass was either unwilted or wilted before ensiling and ensiled with or without application of silage additives (e.g., formic acid and inoculant additives).
Approximately 7,000 kg of each silage was brought to the Agricultural Research Institute of Northern Ireland by covered truck. Each silage was mixed thoroughly in a diet mixer to achieve uniformity and then dispensed into, and stored in, polythene-lined evacuated and sealed wooden boxes of 400-kg capacity according to the procedures described by Pippard et al. (1996). Silages were stored for 2 to 27 d before feeding. There was little deterioration of the silages during storage and their chemical composition and fermentation variables remained relatively constant (Pippard et al., 1996).
Digestibility Trials with Sheep
For each of the 136 silages, nutrient digestibility and urinary energy output were measured in digestibility trials using 72 castrated male sheep (1 yr old) with a mean live weight of 50 kg (45 to 55 kg). These measurements were undertaken in 17 periods of 3-wk duration, with eight silages per period. Before the digestibility trials, the 72 sheep were randomly divided into two groups with 36 sheep per group, which were used in alternate periods. During each period, 32 sheep were used for the evaluation of the eight silages (4 sheep per silage), and the remaining 4 sheep were used for the determination of digestibility of a standard grass hay. The standard grass hay was made from the perennial ryegrass and contained 80.5% DM, 18.1 MJ/kg GE, 8.6% CP, 42.4% ADF, and 70.0% NDF (DM basis). The four sheep were randomly selected for each of the eight silages and the hay. The other group of sheep, when not on experiment, were housed as a single group and offered a standard grass silage ad libitum for 3 wk. This allowed these 36 sheep to adjust their metabolism from the effects of the digestibility trials. The digestibility data obtained from the standard hay were used for the adjustment of effects of animals and periods on the digestibility data from the silages.
During the digestibility trials, the silages and the standard hay were offered as the sole feed to the sheep once daily in the morning at maintenance feeding level, without any supplement (AFRC, 1993). The sheep were housed individually in pens during the first 2 wk and in metabolism crates during the 3rd wk. The crates were designed for separation of feces and urine. Total feces and urine were collected during the final 6 d of each 3-wk period. Feces were weighed daily, stored at 2 to 4°C, and composited for each animal at the end of the 6-d collection period. The fresh feces were then completely mixed and a sample taken for chemical analysis. The daily urine output was collected into 20 mL of 50% (vol/vol) sulfuric acid to ensure ammonia preservation and weighed; a 25% aliquot was retained daily and composited for the 6-d collection period for GE determination.
Measurements
During the preparation of silages for the sheep digestibility study, a representative sample of each silage was retained for the analysis of fermentation variables and nutrient concentration. Each sample was then divided into three portions. One portion (approximately 500 g) of fresh silage was used for analysis of oven DM, toluene DM, ammonia N, Kjeldahl N, and acid-insoluble N using the methods described by Steen (1989), whereas alcohol (ethanol and propanol), VFA, and lactic acid concentration were determined using aqueous extracts of silage (Porter, 1992a). The pH of the silage was estimated according to AOAC (1980) methods, and GE in silage was determined on a fresh silage sample ground in liquid N (Porter, 1992b). The second portion (approximately 500 g) of each silage was dried at 60°C for 48 h and milled for the determination of water-soluble carbohydrates (WSC; Thomas, 1977). The third portion (approximately 500 g) of each silage was dried at 85°C for 18 h and ground to pass a 1-mm screen for the determination of ash, ADF, NDF, acid-insoluble lignin, and ether extract using the methods of AOAC (1980).
The DM, ash, N, GE, NDF, and ADF concentrations in the standard grass hay were determined using the methods as previously described.
The N concentration in feces was analyzed in fresh samples using a Kjeldahl Auto 1030 Analyzer (Tecator, Hoganas, Sweden). A fresh sample of feces was dried at 100°C over 48 h for determination of DM concentration. The dried sample of feces was milled through a 1-mm screen and analyzed for GE, NDF, and ash. The GE concentration in urine was measured in a 10-mL freeze-dried sample, which was contained in a self-sealing polythene (polyethylene) bag of known weight and energy concentration. The methods used for chemical analysis of feces and urine were previously described.
Data Analysis
In the present experiment, digestibilities of DM, OM, CP, GE, and NDF and DOMD in grass silages were measured, whereas ME concentrations in silages were estimated as the difference between DE intake and energy outputs from urine and as methane (predicted). The methane energy output was calculated using the equation of Blaxter and Clapperton (1965).
The digestibility data obtained from the standard hay were used for the adjustment of effects of animals and periods on the digestibility data from the silages. Data were analyzed by analysis of variance according to the changeover design whereby the standard hay was fed to four different animals in each period. The analysis of variance model for each variable included animal, period, and silage effects. Hence, silage means adjusted for animal and period effects were produced.
Correlation coefficients were determined using linear regression between energy or digestibility parameters and nutrient concentration or fermentation variables in silages, [Eq. I].
 | [I] |
 | [II] |
 | [III] |
This linear regression Eq. [I
], together with the quadratic regression Eq. [II], were used to relate each digestibility value to CP, NDF, or ADF concentration in grass silages. The stepwise multiple regression technique was used to develop prediction equations for DE and ME concentrations and digestibility data using nutrient concentration and fermentation variables in silages. The technique automatically selected the best predictors to fit the prediction equations (i.e., Eq. [III]). All analyses were performed using Genstat (6th ed.; Lawes Agricultural Trust, Rothamsted, England, U.K.).
Results
Silage Data and Digestibility Variables
The mean, standard deviation, minimum, and maximum nutrient concentration, and fermentation variable data are presented in Table 1. Large differences were observed in the data. For example, the differences between maximum and minimum data in nutrient concentration were 25.8% DM, 3.7 MJ/kg GE, 13.3% CP, 18.3% ADF, and 24.9% NDF (DM basis). The corresponding data for pH, ammonia N/total N, acetic acid, and lactic acid concentrations were 2.0, 34.0%, 5.8%, and 14.4% (DM basis), respectively.
|
. In accordance with the large differences in nutrient concentrations and fermentation characteristics, there were large ranges in DE and ME concentration and for digestibility response variables. The maximum value was proportionately 0.54 greater than the minimum value for DE concentration and 0.68 for ME concentration. The corresponding values for digestibility data of DM, OM, GE, CP, and NDF and DOMD were 0.46, 0.45, 0.49, 0.76, and 0.64 and 0.45, respectively.
|
The correlation coefficients of linear relationships between the energy concentration or digestibility data and nutrient concentration or fermentation variables in the silages are presented in Table 3. Examples are given in Figure 1, representing, respectively the relationships between CP concentration and CP digestibility, NDF concentration and OM digestibility, lignin concentration and DOMD, and lactic acid/total VFA and energy digestibility. There was no significant relationship of any energy concentration or digestibility variable with DM, WSC, or total alcohol (ethanol and propanol) concentration in the silages. The nutrient concentrations and fermentation variables were related (P < 0.05) to the majority of the digestibility and energy data (Table 3). The relationship of DE, ME or digestibility data with lactic acid/total VFA or concentration of CP, soluble CP, GE, ether extract, lactic acid, or ethanol, if significant, was positive. In contrast, relationships with pH, ammonia N/total N, or concentration of ADF, NDF, lignin, acetic acid, propionic acid, butyric acid, valeric acid, total VFA, or propanol were negative. Generally, relating digestibility, DE or ME data to nutrient concentration in silages produced greater correlation coefficients than those when relating to fermentation variables. Among these relationships, those relating to NDF concentration generally had the highest correlation coefficients, followed by lignin, ADF, soluble CP, ether extract, and CP concentrations in silages. When relating digestibility, DE, or ME data to fermentation variables, the correlation coefficients were relatively high with lactic acid concentration, lactic acid/total VFA, or ammonia N/total N, whereas with other data they were relatively low, although significant (P < 0.05).
|
|
and with CP concentration in Table 5. All relationships were significant (P < 0.001). When relating to ME/GE; DOMD; and digestibility of DM, OM, GE, and NDF, generally, NDF concentration fitted the line better than ADF and CP concentrations, with the former having a greater r2 value. However, when relating to CP digestibility, CP concentration produced a greater r2 value than NDF concentration. These linear equations indicated that increasing ADF or NDF concentration in silages by 0.10 kg/kg of DM decreased the digestibility data by proportionately 0.090 to 0.109 or 0.080 to 0.096, and ME/GE by 0.080 or 0.072. However, digestibility and ME/GE could be increased by proportionately 0.121 to 0.228 and 0.103, respectively, by increasing CP concentration by 0.10 kg/kg of DM.
|
|
). Multiple Prediction Equations for Energy Concentrations and Nutrient Digestibility
The multiple prediction equations for energy parameters and digestibility data in grass silages using nutrient concentration and fermentation variables, developed using the stepwise multiple regression technique, are presented in Tables 6 and 7. All relationships were significant (P < 0.001) and each predictor had a significant effect on the relationships (P < 0.05). Three sets of equations were developed for each parameter (i.e., a, b, and c). The "a" equations ([8a] through [16a]) were developed using all nutrient concentration and fermentation variables available in the present experiment. These equations included soluble N proportion and lignin concentration, which may not be measured in grass silages in normal practice, and consequently these two variables were excluded in the development of the "b" equations ([8b] through [15b]). There was no equation of [16b] for prediction of NDF digestibility because addition of any fermentation variable did not have significant effects on the equation. The "c" equations ([8c] through [16c]) were developed as supplements for the above two sets of equations using only nutrient concentration in grass silages. There were strong relationships between energy/digestibility data and nutrient concentrations and fermentation variables. The R2 values were relatively high in the nine equations in 8a through 16a, ranging from 0.67 for predicting NDF digestibility (Eq. [16a]) to 0.80 for predicting CP digestibility (Eq. [15a]). The R2 values were decreased when soluble N/total N and lignin concentration were excluded (Eq. [8b] through [15b]) and further decreased when using only GE or CP and ash and NDF concentrations as the predictors (Eq. [8c] through [16c]). For prediction of DE/GE, ME/GE, and DE and ME concentrations, GE and NDF concentrations had the major effects on the relationships, whereas the main predictors were CP and NDF concentrations for the prediction of digestibility of DM, OM, CP and NDF, and DOMD.
|
|
Nutrient Concentration and Digestibility
The large range in silage quality and the uniform distribution of silages within the range were the key aspects enabling development of accurate prediction equations for silage energy concentration and digestibility in the present experiment. To achieve this objective, silages (n = 136) used were selected using the preset criteria (pH, ammonia N/total N, DM, and ME concentrations) from thousands of grass silages produced on commercial farms across Northern Ireland. The nutritive values and fermentation characteristics of the silages (Tables 1 and 2) indicated that the above objective was achieved. The silages used were also relatively evenly distributed within the whole range. For example, the CP and OM digestibility and DOMD were evenly plotted against CP, NDF, and lignin concentrations in the silages (Figure 1) from the low to top ranges, respectively.
The grass silages used in the present experiment were within the ranges of clamp silages produced in the United Kingdom in terms of silage quality. For example, toluene-corrected DM, CP, GE, and ME concentrations in clamp grass silages (n = 180) presented in U.K. Tables of Nutritive Value and Chemical Composition of Feedingstuffs (MAFF, 1990) range from 0.175 to 0.515 kg/kg, 0.087 to 0.303 kg/kg of DM, and 14.7 to 21.7 and 7.0 to 13.7 MJ/kg of DM, respectively. The corresponding digestibility data of DM and DOMD range from 0.48 to 0.79 and from 0.491 to 0.782 kg/kg of DM. These ranges are marginally greater than those in the present experiment. The silages used in the present experiment would therefore embrace the majority of clamp grass silages used on commercial farms in the United Kingdom.
Relationship Between Energy/Digestibility Data and Nutrient Concentration and Fermentation Variables in Grass Silages
In the present experiment, strong linear relationships were obtained between silage fermentation characteristics. For example, total VFA concentration was negatively related to lactic acid concentration (P < 0.001) with an r2 value of 0.39 but positively related (P < 0.001) to both pH (r2 = 0.51) and ammonia N/total N (r2 = 0.48). The relationship between pH and ammonia N/total N was positive and even stronger (P < 0.001, r2 = 0.68). It is therefore logical in the present experiment that the digestibility of DM, OM, and GE, as well as DOMD and ME/GE, were all negatively related to pH, ammonia N/total N, and total VFA (or each individual VFA) concentration, whereas they were positively related to lactic acid concentration and lactic acid/total VFA. Although there is little information available on these relationships in the literature, numerous publications have demonstrated the relationships between silage fermentation characteristics and voluntary silage intake. For example, a number of studies have found that silage DM intake decreased with increasing ammonia N/total N in silages (Steen et al., 1995). This measurement accounted for proportionately 0.38 and 0.42 of the variation in the intake of non-formaldehyde-treated silages given to sheep (Wilkins et al., 1971, 1978). Negative relationships between acetic, butyric, or total VFA and silage intake have also been reported (Rook and Gill, 1990). When collinearity was removed, butyric acid still strongly affected intake, indicating that it has an important depressing effect on intake, independent of other components within the silages (Steen et al., 1995).
In the present experiment, DE and ME concentrations; ME/GE; DOMD; and digestibility of DM, OM, CP, GE and NDF were all significantly increased with increasing CP, soluble CP, and ether extract concentrations in silages but significantly decreased with increasing fibrous fractions (NDF, ADF, and lignin). Generally, CP (soluble CP) and fibrous fractions (NDF and lignin) were the most important factors among the nutrient concentrations and fermentation characteristics. In the linear relationships presented in Table 4, increasing 0.10 kg/kg of DM of NDF and ADF decreased DOMD by proportionately 0.080 and 0.090, respectively (Eq. [5a] and [5b]). Similar decreases have been reported in a number of studies. For example, with each increase in 0.10 kg/kg of DM of NDF or ADF concentration, DOMD was reported to be proportionately decreased by 0.062 or 0.072 in clamp grass silages (n = 124; Givens et al., 1989), 0.071 or 0.099 in big bale grass silages (n = 37; Givens et al., 1993a), and 0.067 or 0.090 in fresh grass (n = 173; Givens et al., 1990a). Similar decreases in DOMD with increasing NDF and ADF concentration were also observed by Nousiainen et al. (2003) in clamp grass silages. This decrease consequently leads to a decrease in forage ME concentration. For example, ME concentration in fresh grass (n = 173) was reported to be decreased by 1.33 or 1.70 MJ/kg of DM with increasing NDF or ADF concentrations of 0.10 kg/kg of DM (Givens et al., 1990b).
In contrast, increasing CP concentration in silages increases nutrient digestibility. A previous study has shown that the increase in DOMD was proportionately 0.074 with an increase in CP concentration of 0.10 kg/kg of DM in fresh grass (Givens, 1990a) and 0.171 in clamp grass silages (Nousiainen et al., 2003). The positive relationship between CP digestibility and CP concentration was also found in grass silage (n = 50) by Steg et al. (1990). A similar positive relationship (linear) between DOMD or CP digestibility and CP concentration in grass silages was also noted in the present experiment (Table 5). However, CP digestibility was better fitted to CP concentration in a curvilinear relationship (quadratic; Figure 1a), indicating that increases in CP digestibility would be decreased with increasing CP concentration in silages over a certain point. The asymptote for CP digestibility in the present experiment was at a CP concentration of 0.193 kg/kg of DM. A similar result was also found in the present experiment when relating soluble CP concentration to CP digestibility, with the r2 values (0.67 for linear and 0.74 for quadratic relationship) being greater than those using CP concentration. The asymptote for depression of CP digestibility was 0.115 kg/kg of DM of soluble CP in grass silages.
Three sets of multiple prediction equations were developed in the present experiment to predict DE and ME concentration; ME/GE; DOMD; and digestibility of DM, OM, GE, CP, and NDF using the stepwise multiple regression technique. The predictors used were gradually restricted from all nutrient concentrations and fermentation variables to nutrient concentrations only. The pattern of effects of these selected predictors on the multiple equations were the same as seen in the linear relationship, i.e., the effects of GE, CP, soluble CP proportion, and lactic acid/total VFA on energy and digestibility data were positive, whereas NDF, lignin, and ammonia N/total N were negative. However, DM and ash concentrations when considered alone in the linear regression had no significant effect on DE or ME concentration or each digestibility value but had a negative effect on these multiple equations. In multiple regression, the negative effect of ash on DOMD has also been reported by Givens et al. (1989) in clamp grass silages, by Givens et al. (1990a) in fresh grass, and on OM digestibility by Steg et al. (1990) in clamp silages. Steg et al. (1990) also found that DM concentration of grass silages had a negative effect on OM digestibility using multiple regression equations.
Validation of Multiple Prediction Equations Developed in the Present Experiment Using Published Data
The data used in the present validation were obtained from 12 studies (Beever et al., 1977; Kelly and Thomas, 1978; Berge, 1979; Morgan et al., 1980; Thompson et al., 1981; McLellan and McGinn, 1981, 1983; Givens et al., 1989, 1993a, Givens et al., b; Steg et al., 1990; Moss et al., 1995). The forages used in these studies were grass silages offered alone as a sole feed at or near maintenance feeding level. The animals used were cattle in the studies of Morgan et al. (1980) and Thompson et al. (1981), whereas in the remaining 10 studies, sheep were used. All data (n = 28) were derived from single silages (n = 24), with exception of the studies of Givens et al. (1989, 1993a,b) and Steg et al. (1990), in which the data were mean values from 124, 37, 9, and 50 silages, respectively. Organic matter digestibility and DOMD were determined in all studies, but DE and ME concentrations; ME/GE; and digestibility of DM, GE, and CP were not reported in some of the studies. Only 3 of the 12 experiments reported NDF digestibility data, and therefore NDF digestibility relationships were not validated for the present experiment. Because soluble N proportion and lignin concentration in silages were not reported in the majority of the studies used in the validation, the present validation was only performed for Eq. [8b] through [15b] and [8c] through [15c], as presented in Tables 6 and 7.
The prediction accuracy of relationships was examined using the mean-square prediction error (MSPE) as described by Rook et al. (1990). The MSPE is defined as Eq. [IV] and can be regarded as the sum of three components (Eq. [V]):
 | [IV] |
 | [V] |
where P or A is predicted or actual DE (or ME) concentration, ME/GE, digestibility of DM, OM, GE or CP or DOMD; n is the number of pairs of values of P and A compared; 
or 
is the mean of P or A; SP2 or SA2 is the variance of P or A; b and r are the slope and correlation coefficient, respectively, of the linear regression of P on A. The three components are thus due to mean bias ( – ), line bias (the deviation of the slope), and random variation of the slope. Mean prediction error (MPE), rather than MSPE, was used to describe the prediction accuracy (
).
Results of the present validation are presented in Table 8. Silage DM concentration in the studies used for the present validation was considered on a toluene-corrected basis, whereas in the present experiment it was corrected for both toluene and alcohol. Therefore, the predicted values using the equations developed were adjusted from toluene and alcohol to a toluene-corrected basis using the mean alcohol proportion in the alcohol- and toluene-corrected DM obtained in the present experiment. In general, inclusion of fermentation characteristics (lactic acid/total VFA and ammonia N/total N) as predictors (Eq. [8b] through [14b]) decreased both SE and MPE values in comparison with equations that excluded these two parameters (Eq. [8c] through [14c]). However, for prediction of CP digestibility, the SE and MPE values were lower when fermentation predictors were excluded (Eq. [15c] vs. [15b]). The mean predicted values were relatively close to the mean actual data in all equations ([8b] through [15b] and [8c] through [15c]), with the exception of underprediction of ME concentration by 0.3 MJ/kg of DM with Eq. [9c] and overprediction of CP digestibility by 0.03 kg/kg with Eq. [15b]. Accordingly, a large error derived from the bias (predicted minus actual) as a proportion of total MSPE was derived in Eq. [11c], [15b], and [15c], but in other equations this error was relatively small. The prediction of energy variables (Eq. [8b] through [11b] and [8c] through [11c]) produced a relatively small error of line (slope) as a proportion of total MSPE, whereas this parameter was relatively large in the prediction of digestibility of DM, OM, and CP and DOMD (Eq. [12b] through [15b] and [12c] through [15c]). As a result, the error derived from the random variation was very high in the former eight equations for prediction of energy variables and relatively small in the latter eight equations for the prediction of digestibility data.
|
for DE and ME concentration and DE/GE and ME/GE, and Figure 3 for digestibility of DM, OM, and CP and DOMD. The majority of residual plots for prediction of CP digestibility were positive (overpredicted; Eq. [15b] and [15c]), whereas relatively more plots for prediction of ME concentration and ME/GE were negative (underpredicted; Eq. [9b], [9c], [11b], and [11c]). However, the plots for prediction of other variables (DE concentration; digestibility of GE, DM, and OM; and DOMD) were distributed around zero. The SD values for the residual differences (predicted minus actual data) were in accordance with SE and MPE values presented in Table 8 (i.e., the SD values in prediction equations including fermentation variables as predictors were generally smaller than those excluding fermentation variables).
|
|
Implications
Grass silage is a major feed for ruminant animals across the world and its nutritive value varies greatly. As reported in the present experiment, there are significant relationships between digestible or metabolizable energy concentration or nutrient digestibility and nutrient concentrations or fermentation characteristics in grass silages, especially concentrations of crude protein and neutral detergent fiber. Based on these relationships, three sets of multiple prediction equations for nutritive values in grass silages have been developed and validated using published grass silage data. These equations can be used to calculate digestible and metabolizable energy concentrations in the grass silage using routinely determined silage data (dry matter, ash, crude protein, gross energy, neutral detergent fiber, lactic acid, total volatile fatty acid concentrations, and ammonia nitrogen).
Footnotes
1 The authors thank their colleagues at the Agric. Res. Inst. of Northern Ireland for access to the data used in the present study. 
3 Also a member of staff of the Dept. of Agric. and Rural Development of Northern Ireland and the Queen’s Univ. of Belfast, Belfast, Northern Ireland BT9 5PX.
2 Correspondence—phone: 44 (0)28 9268 2484; fax: 44 (0) 28 9268 9594; e-mail: tianhai.yan{at}dardni.gov.uk.
Received for publication August 15, 2003. Accepted for publication February 2, 2004.
Literature Cited
AFRC. 1993. Energy and protein requirements of ruminants. Agricultural and Food Research Council. CAB International, Wallingford, Oxon, U.K.
Agnew, R. E., and T. Yan. 2000. The impact of recent research on energy feeding systems for dairy cattle. Livest. Prod. Sci. 66:197–215.[Medline]
AOAC. 1980. Official Methods of Analysis. Assoc. Offic. Anal. Chem. Washington, DC.
Barber, G. D., D. I. Givens, M. S. Kridis, N. W. Officer, and I. Murray. 1990. Prediction of the organic matter digestibility of grass silage. Anim. Feed Sci. Technol. 28:115–128.
Beever, D. E., D. J. Thomson, S. B. Cammell, and D. G. Harrison. 1977. The digestion by sheep of silages made with and without the addition of formaldehyde. J. Agric. Sci. (Cambridge) 88:61–70.
Berge, P. 1979. The effect of the proportion of supplement and the level of feeding on the digestibility of grass and red clover silages by sheep. M.S. Thesis, Univ. Reading, Reading, England.
Blaxter, K., and J. L. Clapperton. 1965. Prediction of the amount of methane produced by ruminants. Br. J. Nutr. 19:511–522.[Medline]
Givens, D. I., J. M. Everington, and A. H. Adamson. 1989. The digestibility and metabolisable energy content of grass silage and their prediction from laboratory measurements. Anim. Feed Sci. Technol. 24:27–43.
Givens, D. I., J. M. Everington, and A. H. Adamson. 1990a. The nutritive value of spring-grown herbage produced on farms throughout England and Wales over 4 years. 2. The prediction of apparent digestibility in vivo from various laboratory measurements. Anim. Feed Sci. Technol. 27:173–184.
Givens, D. I., J. M. Everington, and A. H. Adamson. 1990b. The nutritive value of spring-grown herbage produced on farms throughout England and Wales over 4 years. 3. The prediction of energy values from various laboratory measurements. Anim. Feed Sci. Technol. 27:185–196.
Givens, D. I., A. R. Moss, and A. H. Adamson. 1993a. Prediction of the digestibility and energy value of grass silage conserved in big bales. Anim. Feed Sci. Technol. 41:297–312.
Givens, D. I., A. R. Moss, and A. H. Adamson. 1993b. The digestibility and energy value of badly preserved grass silages. Anim. Feed Sci. Technol. 42:97–107.
Kelly, N. C., and P. C. Thomas. 1978. The nutritive value of silages. Energy metabolism in sheep receiving diets of grass silage or grass silage and barley. Br. J. Nutr. 40:205–219.[Medline]
MAFF. 1990. UK tables of nutritive value and chemical composition of feedingstuffs. Ministry of Agriculture, Fisheries and Food. D. I. Givens and A. R. Moss, ed., Rowett Research Services Ltd., Aberdeen, U.K.
McLellan, A. R., and R. A. McGinn. 1981. Calorimetric investigation of the nutritive value of formic acid- and formaldehyde-formic acid-treated grass silages. Anim. Feed Sci. Technol. 6:245–256.
McLellan, A. R., and R. A. McGinn. 1983. Energy and nitrogen balance, intake and growth rate of lambs fed on unwilted grass silages treated with a formaldehyde-acid mixture or untreated. Anim. Feed Sci. Technol. 9:107–119.
Morgan, C. A., R. A. Edwards, and P. McDonald. 1980. Intake and metabolism studies with fresh and wilted silages. J. Agric. Sci. (Camb.) 94:287–298.
Moss, A. R., D. I. Givens, and P. C. Garnsworthy. 1995. The effect of supplementing grass silage with barley on digestibility, in sacco degradability, rumen fermentation and methane production in sheep at two levels of intake. Anim. Feed Sci. Technol. 55:9–33.
Nousiainen, J., M. Rinne, M. Hellämäki and P. Huhtanen. 2003. Prediction of the digestibility of the primary growth of the grass silages harvested at different stages of maturity from chemical composition and pepsin-cellulase solubility. Anim. Feed Sci. Technol. 103:97–111.
Pippard, C. J., M. G. Porter, R. W. J. Steen, F. J. Gordon, C. S. Mayne, R. E. Poots, E. F. Unsworth, and D. J. Kilpatrick. 1996. A method for obtaining and storing uniform silage for feeding experiments. Anim. Feed Sci. Technol. 57:87–95.
Porter, M. G. 1992a. Comparison of five methods for the determination of lactic acid in silage. Pages 123–124 in Br. Grass. Soc. 3rd Res. Conf., Antrim, U.K.
Porter, M. G. 1992b. Comparison of sample preparation methods for the determination of the gross energy concentration of fresh silage. Anim. Feed Sci. Technol. 37:201–208.
Rook, A. J., M. S. Dhanoa, and M. Gill. 1990. Prediction of the voluntary intake of grass silages by beef cattle. 3. Precision of alternative prediction models. Anim. Prod. 50:455–466.
Rook, A. J., and M. Gill. 1990. Prediction of the voluntary intake of grass silages by beef cattle. I. Linear regression analyses. Anim. Prod. 50:425–438.
Steen, R. W. J. 1989. A comparison of soya-bean, sunflower and fish meals as protein supplements for yearling cattle offered grass silage-based diets. Anim. Prod. 48:81–89.
Steen, R. W. J., F. J. Gordon, C. S. Mayne, R. E. Poots, D. J. Kilpatrick, E. F. Unsworth, R. J. Barnes, M. G. Porter, and C. J. Pippard. 1995. Prediction of the intake of grass silage by cattle. Pages 67–89 in Recent Advances in Animal Nutrition. P. C. Garnsworthy and D. J. A. Cole, ed. Nottingham Univ. Press, Nottingham, U.K.
Steg, A., S. F. Spoelstra, J. M. van der Meer, and V. A. Hindle. 1990. Digestibility of grass silage. Neth. J. Agric. Sci. 38:407–422.
Thomas, T. A. 1977. An automated procedure for the determination of soluble carbohydrates in herbage. J. Sci. Food Agric. 28:639–642.
Thompson, D. J., D. E. Beever, C. R. Lonsdale, M. J. Haines, S. B. Cammell, and A. R. Austin. 1981. The digestion by cattle of grass silage made with formic acid and formic acid-formaldehyde. Br. J. Nutr. 46:193–207.[Medline]
Wilkins, R. J., J. S. Fenlon, J. E. Cook, and R. F. Wilson. 1978. A further analysis of relationships between silage composition and voluntary intake by sheep. Pages 34–35 in Proc. 5th Silage Conf., Ayr, U.K.
Wilkins, R., K. J. Hutchinson, R. F. Wilson, and C. E. Harris. 1971. The voluntary intake of silage by sheep. 1. Interrelationships between silage composition and intake. J. Agric. Sci. (Camb.) 77:531–537.
This article has been cited by other articles:
|
|
 |
|
  J. C. Burns ASAS Centennial Paper: Utilization of pasture and forages by ruminants: A historical perspective J Anim Sci, December 1, 2008; 86(12): 3647 - 3663. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Boguhn, H. Kluth, and M. Rodehutscord Effect of Total Mixed Ration Composition on Amino Acid Profiles of Different Fractions of Ruminal Microbes In Vitro J Dairy Sci, May 1, 2006; 89(5): 1592 - 1603. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Yan and R. E. Agnew Prediction of nutritive values in grass silages: II. Degradability of nitrogen and dry matter using digestibility, chemical composition, and fermentation data J Anim Sci, May 1, 2004; 82(5): 1380 - 1391. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
