Comparison of diets based on grass silage , hay or oat straw supplemented with four levels of concentrates in the feeding of growing Ayrshire bulls

Aronen, I. 1 , Lampila, M. 1 & Hepola, H. 2 1994. Comparison of diets based on grass silage, hay or oat straw supplemented with four levels of concentrates in the feeding of growing Ayrshire bulls. Agricultural Science in Finland 3: 15-26. ('Agricultural Research Centre of Finland, Institute of Animal Production, FIN-31600 Jokioinen, Finland and department of Animal Science, P.O. Box 28, FIN-00014 University of Helsinki, Finland. Present address: Ilmo Aronen, Rehuraisio Ltd, P.O. Box 101, FIN-21201 Raisio, Finland.)


Introduction
Livestock production in Finland is largely based on grass silage feeding.On controlled herds, 32% of the total feed energy for growing cattle is derived from grass silage, only 6% from hay and different types of straw, and 36% from feed grain (Lätti   1991).
There are considerable differences in both the energy and protein content of grass silage, hay and straw.In general, grass silage has the highest and straw the lowest content of digestible energy and protein.Therefore, various levels of concentrates with a different protein content are needed to supplement these diets.The effects of different levels of concentrate supply have been studied in numer- ous experiments, with diets based on grass silage (e.g., Suomi et al. 1975, 1977, Varvikko and  Lampila 1984, Martinsson 1990, Aronen et al.  1992), hay (e.g., Lampila et al. 1987) or straw (e.g., Aronen et al. 1987, Lampila et al. 1987),but only a few reports are available in which different levels of concentrates with different roughages have been studied in the same experiment (e.g., Lampila et al. 1987).
The aim of the present experiment was to com- pare the feed value of grass silage, hay and oat straw as roughages for growing cattle supplemented by four different levels of concentrates.Some preliminary results have been presented ear- lier by Lampila (1988) and Lampila el al. (1988).

Material and methods
This study comprises three production experiments, conducted according to identical arrangements during three consecutive years.

Animals and their feeding
Each experiment included 48 Ayrshire bulls of an average initial age of 185 days and live weight of 186 kg.Four of the bulls were lost during the ex- periments for reasons not related to the treatments.
The treatments were arranged according to 3 * 4 incomplete factorial design.The roughages were fed ad libitum and different amounts of concentrates were fed as follows: Direct-cut grass from a timothy-cocksfoot (Phleum pratense-Dactylis glomerata) and from a cocksfoot-timothy sward was ensiled in tower silos with a formic acid-based additive (AIV II; 80% (w/w) formic acid, 2% ortho-phosphoric acid) ap- plied at the rate of 51/t.The silages were of the first and second cut.
The hay used in this study was field dried and baled cocksfoot-timothy and meadow fescue-timo- thy (Festuca pratensis -Phleum pratense).The straw was field-dried and baled oat straw.
In grass silage-based diets, barley was fed as a concentrate, but in hay-and straw-based diets some of the barley was replaced by rapeseed meal and urea to increase and balance the nitrogen content of the diet.The animals were fed once daily and the feed intake was recorded individually.In feeding groups with a daily concentrate allowance higher than 3 kg, the concentrates were fed in steadily growing quantities, so that the intended allowance was met during the course of the experiment.A commercial mineral mixture was fed at the rate of 150 g/d and a vitamin mixture at the rate of 50 g/week.Furthermore, an injection of fat-soluble vitamins was given to the bulls on straw feeding every 10 weeks.Water was freely available.The bulls were weighed every 14 days.

Analytical methods
Feed samples were taken at every feeding and pooled for periods of two weeks for roughages and four weeks for concentrates.Proximate feed analyses were carried out conventionally.The correction for volatile losses in grass silage was made accord- ing to Huida et al. (1986).Fermentation quality of the grass silages was analyzed as described by Vanhatalo et al. (1992).Apparent in vivo digestibility of the roughages and barley was measured in wethers, and table values (Salo et al. 1990) were applied for rapeseed meal.Metabolizable energy values of the feeds were calculated according to MAFF (1975) and values for net energy according to Salo et al. (1990).
Ruminal N degradability of the feeds was deter- mined with three Ayrshire heifers with the feed samples originally used for proximate feed analysis.The nylon bag technique was used as described by Aronen (1990).Degradability ofcmde protein (DEG) was calculated according to ORSKOV and   McDonald (1979) using a rumen outflow rate (k) of 8%/h as suggested by Hvelplund and  Madsen (1990).No correction for possible small particle losses from the bags was made for concen- trates.However, a correction for microbial contamination in grass silage and hay was made ac- cording to Lindberg (1988) and Michalet- Doreau and Ould-Bah (1989), and a mean value of these two methods was adopted.DEG for straw was calculated according to Lindberg (1988).AAT (amino acids absorbed in the small intestine) and PBV (protein balance in the ramen) were cal- culated according to Madsen (1985).The digestibility of undegraded feed protein was assumed to be 82%.

Carcass evaluation
Carcass weight was estimated after removal of the skin, head, feet, tail, internal organs, and kidney knob and channel fat.Fat in the abdominal cavity was calculated and carcass dissectible fat obtained as described by Olsson and Lindberg (1985).Carcass quality and fatness were graded visually using the carcass classification scheme employed in slaughterhouses in Finland.The gastro-intestinal tract was weighed immediately after slaughter and again after it had been emptied.

Statistical analysis
Due to the incompleteness of the factorial arrangements (the levels of concentrates used with differ- ent roughages were not equal), the effect ofconcen- trate supplementation was analyzed separately within each roughage by an analysis of variance.In addition, the experiment year was included in the model.The effect of concentrate level was further partitioned, using orthogonal contrasts, into linear and quadratic effects.The data related to feed intake and animal performance are given as least squares means.

Results and discussion
Feeds and feed intake Table 1 shows the average chemical composition and feed value of the experimental feeds as simple means.Both the nutritional and fermentation quality of the grass silages was good.No butyric acid was detected and the proportion of ammonia nitro- gen of total nitrogen was small.The digestibility and thereby the feed value of the silages was also good.The hay had an average energy value and the energy content of the straw was relatively high.The concentrate intake in each feeding group was as intended.Regardless of the type of roughage, increase in concentrate intake was found to decrease the roughage intake (Tables 2a-2c).How- ever, total DM intake increased linearly.
On grass silage feeding, the first kg of concen- trates decreased grass silage intake by 0.25 kg DM/kg of concentrate DM, the second by 0.70 kg and the third by 0.56 kg.Intake of hay decreased by 0.84 kg DM/kg of concentrate DM, when concen- trate intake was increased from 1 kg to 2 kg.The third and the fourth kg ofconcentrates reduced hay intake by 0.51 kg and 0.73 kg DM/kg of concen- trate DM, respectively.On straw feeding the de- pression in roughage intake was 0.34,0.67and0.49 kg DM/kg of concentrate DM, respectively.On an average, the depression in roughage intake on grass silage feeding was 0.50 kg DM/kg of concentrate DM, on hay feeding 0.69 kg DM and on straw feeding 0.50 kg DM.
The decrease in roughage intake is usually larger with good quality roughages than with roughages ofpoor quality (BINES 1985).Indeed, MARTINSSON  (1990) reported that per kg of concentrate DM fed, the depression in silage DM intake was greater with an early-cut (0.76 kg) silage than with a late-cut one (0.64 kg).Aronen et al. (1992) reported a substitu- tionrate of 0.63 kg silage DM per kg ofconcentrate DM.In that experiment, the digestibility of organic matter (OM) was on average 76.3%, whereas it was 73.3% in the present experiment.
Due to the varying composition of the concen- trates, it is difficult to interpret the substitution rate with hayor straw-based diets.It has been shown Table 1.Average chemical composition (%), organic matter digestibility (%) and feed value of the experi- mental feeds and fermentation quality of grass silage.In total N (g/kg) Soluble N 511 Ammonium N 40 DEG without correction for microbial contamination in grass silage and hay was 72 and 56, respectively.NFE, nitrogen free extracts; FU, feed unit; ME, metabolizable energy; MJ, megajoule; DCP, digestible crude protein; AAT, amino acids absorbed in the small intestine; PBV, protein balance in the rumen; DEG, degradability of feed protein; N, nitrogen.that protein feeds with differing degradability may have a different effect on roughage intake (e.g., Aronen 1990).
The increase in concentrate feeding level re- sulted in an increase both in energy and protein intake (Tables 2a-2c).The increase in protein intake with animals fed grass silage was relatively larger when the protein intake was calculated as AAT than when calculated as crude protein or di- gestible crude protein.The calculated PBV was clearly higher for grass silage diets than for hay or straw diets.In other words, there was a large excess of rumen degradable protein in grass silage diets.However, increasing the level of concentrates re- duced this excess.At the same time, LWG in-creased (Table 3a), thereby increasing the protein deposition in the tissues.Thus, raising the concen- trate level improved the nitrogen utilization on grass silage feeding.

Animal performance
Regardless of the roughage, the increase in concen- trate intake resulted in a faster rate of daily live weight gain (LWG) (Figure 1).This effect was linear.Raising the amount of concentrates had a similar effect on carcass gain (Tables 3a-3c).These findings are in accordance with results reported by Suomi et al. (1975) and Aronen et al. (1987), who, Table 2a.Effect of concentrate supplementation on feed intake and estimated nutrient consumption in growing cattle on grass silage feeding. in their studies, included increasing amounts of barley in grass silage-based diets or in diets based on oat straw, respectively.A rather good LWG was obtained with grass silage alone.However, the results clearly show that the total growth potential of the Ayrshire bulls could not be utilized by feeding the bulls with grass silage alone or grass silage with small (1 to 2 kg/d) amounts of concentrates.

Effect of
The differences in carcass gain between the dif- ferent roughages were clear.To obtain an equal carcass gain, 2 kg more of concentrates was needed on hay feeding than on grass silage feeding.Fur- thermore, on straw feeding, 4 kg of concentrates SEM, standard error of mean; NS, statistically non-significant; *, p < 0.05; **, p < 0.01; ***, p < 0.001; L, linear effect; Q, cubic effect; DM, dry matter; FU, feed unit; CP, crude protein; DCP, digestible crude protein; AAT, amino acids absorbed in the small intestine; PBV, protein balance in the rumen.
was not enough to obtain comparable carcass gain as with grass silage alone.When the calculated energy intake in fattening feed units was plotted against daily carcass gain (Figure 2), the relative value of the differentrough- ages could be compared.It was shown that, in order to obtain a similar carcass gain, more energy was needed on straw than on hay feeding and that more energy was needed on hay than on grass silage feeding.This observation may imply two things  3a.Effect of concentrate supplementation on the rate of live weight gain (LWG), carcass gain (CG) and feed conversion on grass silage feeding.SEM, standard error of mean; NS, statistically non-significant; *, p < 0.05; **, p < 0.01; *•*, p < 0.001; L, linear effect; Q, cubic effect; FU, feed unit.

Effect of
about the finnish energy system.Firstly, it is possible that the system overestimates the feed value of hay and straw compared to grass silage and that the feed value of straw is overestimated compared to hay.Secondly, it is possible that, compared to bar- ley, the feed value ofroughages may be underesti- mated.These findings were already reported by Lampila and Micordia (1990), and a working group has been established to re-evaluate the present energy system used in Finland.Due to experimental arrangements, the concen- trate level and weight at slaughter are confounding.Therefore, the resultsrelated to feed conversion and carcass composition should be interpreted with caution.
Regardless of the type of roughage, increasing the intake of concentrate was found to impair feed conversion in feed units per kg of daily live weight gain (Tables  3a-3c).This observation is in line with the results reported by Suomi et al. (1975) and Aronen et al. (1992).However, ithas to be pointed out that because the duration of the feeding experiment was the same for each feeding group, the animals receiving more concentrates were rela- tively more matured due to a higher live weight and faster LWG, and they inevitably contained more fat.According to GEAY and MtCOL (1989), the higher the LWG and energy intake, the more rapidly fat is deposited in the carcass.Results reported by Andersen et al. (1984) and Bailey (1989)  support this concept.This, together with the higher live weight at slaughter and concomitant higher maintenance requirement, could explain the impaired feed conversion with high concentrate diets.
When feed conversion was calculated in feed units per kg of carcass gain, the results were less consistent.On grass silage feeding, the increase in concentrate intake at lower levels of concentrate supply had a positive effect on feed conversion, but the poorest feed conversion was measured at the highest concentrate level.On hay feeding, increas- ing the level ofconcentrates impaired feed conver- sion, whereas the opposite was true on straw feed- ing.
An increase in concentrate feeding level had a positive effect on dressing percentage within each roughage (Tables 4a-4c).The improvement in dressing percentage seemed to be curvilinear, at least in grass silage diets.Suomi et al. (1975), Varvikko and Lampila (1984) and Martinsson  (1990) also reported an improved dressing percentage with an increased concentrate feeding level.
Increasing the level of concentrates in the diet improved the carcass quality grade.However, on grass silage feeding the best score was recorded when the second highest level of concentrates was applied.More subcutaneous fat was deposited when the level of concentrates was raised, as indi-Fig.2. Average daily carcass gain and energy intake of the bulls.Table 4a.Effect of concentrate supplementation on slaughter data on grass silage feeding.4a-4c).This result is in agreement with the findings reported by Lampila et al. (1987) with various roughages, Martinsson (1990) and Aronen et al.  (1992) with grass silage-based diets, and Aronen et al. (1987) with straw-based diets.

Effect of
The amount of fat in the abdominal cavity in-creased when the level of concentrates was increased (Tables 4a-4c).There was a close relation between this fat and carcass weight.This finding is in accordance with the observation that growing cattle with a high live weight deposit more fat than protein (Geay and MICOL 1989).On the other hand, it must be pointed out that in the present experiment, the animals having a higher carcass SEM, standard error of mean; NS, statistically non-significant; *, p < 0.05; **, p < 0.01; ***, p < 0.001; L, linear effect; Q, cubic effect; Quality grade: 5 = skinny -10 = very muscular.Fatness grade: 0 = leanest (T) -4 = fatty carcass (R).
Fat in abdominal cavity: The sum of kidney knob and channel fat, omentum and mesentric fat.
weight at slaughter also had a higher LWG.Indeed, Andersen et al. (1984) observed that to obtain a given carcass weight, a higher level offeeding (and a concomitant increase in LWG) increased the rela- tive amount of fat.
The changes in the amount of fat in the abdomi- nal cavity and of carcass dissectible fat within each roughage were linear.However, when the data are studied as a whole, it can be seen that the changes were curvilinear and that the relatively highest fat accretion was recorded with the highest carcass weights (Figure 3).However, as stated earlier, it is not possible to distinguish whether the above dif- ferences in carcass composition were caused by different carcass weights at slaughter or by differ- ences in LWG or, more likely, by both.
In all diets, the increase in concentrate feeding level was found to reduce the contents of the gastro- intestinal tract (Tables 4a-4c).Similarly, there were large differences in the contents of the gastro-intestinal tract between the different types of roughages.Analogously, Otto et al. (1980) found straw feed- ing to increase the contents of the gastro-intestinal tract when compared to grass silage feeding.This suggests that live weight and LWG are not proper measures to use when diets with different fill char- acteristics are compared.

Conclusions
The results of the present experiment show that grass silage is of superior nutritional value com- pared to hay or straw.To obtain an equal carcass gain, 2 kg more of concentrates was needed on hay feeding than on grass silage feeding.On straw feed- ing, 4 kg of concentrates was not enough to obtain a comparable carcass gain as with grass silage alone.However, the response in carcass gain to concentrate supplementation on grass silage feed- ing indicated that the growth potential of the bulls receiving silage alone or silage with small amounts (1 or 2 kg/day) of concentrates was not completely utilized.

Fig. 1 .
Fig. 1.Average daily live weight gain of the bulls.

Fig. 3 .
Fig.3.Relationship between carcass weight and amount of fat in the ab- dominal cacity.

Table 2b .
Effect of concentrate supplementation on feed intake and estimated nutrient consumption in growing cattle on hay feeding.

Table 2c .
Effect of concentrate supplementation on feed intake and estimated nutrient consumption in growing cattle on straw feeding.

Table 3b .
Effect of concentrate supplementation on the rate of live weight gain (LWG), carcass gain (CG) and feed conversion on hay feeding.
* NS Table3c.Effect of concentrate supplementation on the rate of live weight gain (LWG), carcass gain (CG) and feed conversion on straw feeding.*** *

Table 4b .
Effect of concentrate supplementation on slaughter data on hay feeding.

Table 4c .
Effect of concentrate supplementation on slaughter data on straw feeding.