Effects of variety , soil type and nitrogen fertilizer supply on the nutritive value of barley for growing pigs

Jarmo Valaja Agricultural Research Centre ofFinland, Swine Research Station, Tervamäentie 179, FIN-05840 Hyvinkää, Finland. Current address: Agricultural Research Centre ofFinland, Institute ofAnimalProduction, FIN-31600 Jokioinen, Finland, e-mail: Jarmo, valaja @mtt.fi Kaija Suomi, Timo Alaviuhkola Agricultural Research Centre ofFinland, Swine Research Station, Tervamäentie 179, FIN-05840 Hyvinkää, Finland Timo Mela Agricultural Research Centre ofFinland. Institute ofCrop and Soil Science, FIN-31600, Jokioinen, Finland


Introduction
Barley is the most commonly grown cereal crop in Finland and is mainly used as pig feed.It provides most of the energy and also about half of the protein in pig diets.However, the quality of barley protein is not optimal for pigs since the content ofessential amino acids, lysine and thre- onine in particular, is very low.The main com- ponent affecting the utilization of the nutrients in barley is fibre (Taverner andFarrell 1981, Jacyno 1995).An increased fibre content lowers the energy content and can also impair the di- gestibility of protein and amino acids (Taverner andFarrell 1981, Bach Knudsen andEggum 1984).
Many factors, most notably climate, soil, variety and fertilizer, can influence the grain yield, composition and nutritive value of bar- ley.Earlier experiments revealed differences between barley varieties in digestibility and utilization by pigs (Just et al. 1983 a, b).Fuller et al. (1989) found higher apparent ileal ami- no acid digestibilities in high protein than in low protein barley varieties.Nitrogen (N) fer- tilizer has increased the crude protein content of grain but lowered the lysine content in the protein (Thomke 1970(Thomke , 1976)).In some experiments soil type and cultivation location have also caused variations in the nutritive value of barley (Just et al. 1983b, Thomke 1972).The effects of all three factors on the nutritive value have, however, only rarely been studied in the same experiment (Just et al. 1983 a, b).Knowledge of differences in the protein and energy value of barley may improve the accu- racy of diet formulation and possibly also af- fect the amount of supplementary protein con- centrates needed in pig diets.
The objective of the present experiment was to investigate the effects of variety, N fertilizer supply and soil type on the nutritive value of barley for growing pigs.The nutritive value of barley samples was assessed by their in vitro di- gestibility and the growth performance of the pigs.Some observations of the utilization of the protein by pigs related to N intake were also made.

Barley batches
Three barley varieties, the two-rowed variety Kustaa and the six-rowed varieties Arra and Rokko, were chosen for the experiment.They were grown on either mould or clay soil.N fertilizer was applied at two levels to both soil types; 43 or 71 kg N/ha to mould soil and 76.5 or 110 kg N/ha to clay soil, to represent a low or normal level of N supply for each soil type.Twelve bar- ley batches were produced at the Institute ofCrop and Soil Science of the Agricultural Research Centre in Jokioinen,southern Finland (60°49'N and 23°30'E), in growing season 1990.
Animals, diets and experimental procedure A performance trial was conducted on 240 Yorkshire and Landrace pigs.The pigs, with an aver- age weight of 25.0 kg (SE 0.33), were assigned by sex, weight and litter origin to one of 12 treat- ments based on the 12barley batches.Two pigs were placed in each pen, gilts and castrates be- ing kept apart.Each diet was tested on five pens of gilts and five pens of castrates.The treatments were arranged 3X2X2 factorially to test the effects of variety, N fertilizer supply and soil type.
Each barley (845 g/kg) was supplemented with a constant level of soya bean meal (120 g/kg) and mineral and vitamin mixture (35 g/kg).The barleys were ground in a hammer mill to pass a 3.5mm mesh.The protein and amino acid supply of the pigs varied owing to the variation in the protein and amino acid con- tent of the barley batches.The digestible crude protein (CP) and lysine contents in the diets ranged from 121 to 144 g/kg and from 5.2 to 5.5 g/kg, respectively.The protein and lysine contents in the diets were kept below the level recommended in the nutritional requirements (Salo et al. 1990) in order to maximizethe amount of barley in the diets and to reveal differences in the quality of protein in barley for pigs.
The pigs were housed in partially slatted, concrete-floored pens providing free access to water.They were weighed at 2-week intervals, and their feed consumption was determined dai- ly pen by pen.The pigs were fed twice a day on a restricted scale in relation to age (1.4 to 2.9 kg/ pig/day); the daily allowance was increased by 0.2 kg/week at the beginning, and by 0.1 kg/week after the eighth experimental week.Before feed- ing the diets were mixed with water.The pigs were slaughtered when they reached a weight of 95 kg, and their carcass weight was recorded 24 h after slaughter.The carcass quality of the pigs was determined as described by Valaja et al. (1993).

Chemical analyses
A standard feed analysis was conducted on the barley samples and soya bean meal (AOAC 1984).The dry matter (DM) of the feeds and in vitro residues was determined after 24 h at 105°C.The DM content of the diets was determined at 2-week intervals.The CP content of the feeds and in vitro residues were analysed as 6.25 * Kjeldahl N. Ether extract (EE) was ana- lysed after acid hydrolysis.The amino acids of the barley samples and soya bean meal were de- termined by gas chromatography after hydrolysis with 6 N HCI at 110°C for 20 h (Näsi and Huida 1982).The barley samples were also ana- lysed for neutral detergent fibre (NDF) content by the method of Robertson and Van Soest (1981), for total dietary fibre content by the method of Asp et al. (1983) and for (3-glucans content by the method of McClear and Glennie- Holmes (1985).In vitro assay was performed on the barley samples by the method of Boisen and Fernandez (1991), a two-step procedure compris-ing a 6-hour incubation in pepsine solution fol- lowed by an 18-hour incubation in pancreatin solution.Instead ofcontinuous magnetic stirring, handmixing every half-hour was used during the first incubation step.

Calculations and statistical analysis
The net energy content and feed unit values of the barley samples were calculated with the Finnish energy evaluation system for pigs (Tuori et al. 1995) using constant digestibility coefficients.The data were analysed by the GLM procedure of SAS (1985).The four-way analysis of vari- ance model used to analyse the data was: Y..., =ii+V.+ S. +N.+ SE + VS, +VN, where Y. jk|mn is the dependent variable, pis the overall mean, V. is the effect of variety, S. is the effect of soil type and N k is the effect of N fertilizer; VS.., VN, and SR, are the interactions ' jk' jl kl between the factors; and SE is the effect ofsex.m e is a normal distributed random variable.The N fertilizer supply was pooled to two levels, low and normal, for the statistical analysis as it was assumed that the total available N from soil and fertilizer was approximately the same in both soils.The difference between variety means was tested with Tukey's test.The relations between in vitro digestibilities and chemical constituents were calculated by linear regression equations.
In all analyses the data on one pen were treated as an experimental unit.

Chemical composition and in vitro digestibility
The CP and lysine contents of the barley batches ranged from 120 to 161 g/kg DM and from 27 to 35 g/160 g N, respectively (Table 1).Arra con- tained, on average, 21 g/kg more CP than did Pokko or Kustaa.The lysine content in the pro- tein was fairly similar in all varieties, whereas the lysine content in DM was highest in Arra (4.0, 4.3 and 4.0 g/kg DM in Kustaa, Arra and Pokko, respectively).Pokko contained slightly less Pglucans than did Kustaa or Arra.The N fertiliz- er supply slightly increased the CP content of the barley samples (133 vs. 141 g/kg) but con- comitantly decreased the amount of lysine in the protein (31 vs. 29 g/!6og N).
The in vitro digestibility of DM and N in bar- ley batches ranged from 65.9% to 71.9% and from 85.1 % to 88.6%, respectively, the variation being higher in DM digestibility.The mean di- gestibility of DM in Kustaa, Arra and Pokko was 67.3%, 71.0% and 69.8%, respectively, being highest in Arra.The mean digestibility of N in Kustaa, Arra and Pokko was 87.3%, 88.1% and 85.9%, being also highest in Arra (Table 1).The following regression equation was calculated for the in vitro digestibility of N (Y) as dependent and the CP, NDF and (i-glucan contents as inde- pendent variables: Y = 86.38 (SE 3.784) + 0.49 (SE 0.093) X CP -0.40 (SE 0.117) X NDF +0.76 (SE 0.332) X p-glucan (P<0.001;r 2 =0.91).
No such equation could be calculated for the in vitro digestibility of DM.
Performance and carcass quality of the pigs The performance results are presented as the main factors in Tables 2 and 3 since the interac- Vol.6 (1997): 295-303.
tions were insignificant.The pigs on diets com- posed ofArra grew faster than did those on diets composed of Kustaa (P<0.05); the growth rate of pigs on diets composed of Pokko was inter- mediate.The pigs on Arra-based diets consumed less feed as DM than did the pigs on Kustaa- based diets (P<0.05).However, the N intake of the pigs on Arra-based diets was the highest among the varieties (P<0.05).The feed conver- sion ratio (FCR) of the pigs on diets composed of Arra was better than that of the pigs on diets composed ofKustaa or Pokko (P<0.05).A higher N fertilizer supply ofbarley improved the FCR of the pigs (P<0.01),but also increased their N intake (P<0.01).The pigs fed barley grown on mould had a better FCR than did those fed barley grown on clay soil (P<0.05).
The carcass lean content of the pigs on Arra- based diets was higher than that of the pigs on Kustaa-based diets (P<0.05) and intermediate for the pigs on Pokko-based diets.The N fertilizer supply of barley increased the content of car- cass lean (P<0.05) and lean in valuable parts (P<0.05).
The castrated male pigs gained weight faster (P<0.001) and had a better FCR (PcO.OOl) than did the gilts.In contrast, the gilts produced leaner carcasses than did the castrated males.The differences were highly significant in all carcass quality measurements.

Discussion
Chemical composition N fertilizer supply increased the CP content of the barley samples and decreased the content of lysine in the protein.A similar trend in CP and lysine contents was observed in the experiments

53.71
Significance:NS=non-significant, *=P<o.os,**=P<o.ol.S=soil type.N=nitrogen supply.SEM=standard error of means.LS=least square.FCR=feed conversion ratio.FU=feed unit.NE=net energy.MJ=mega joule. of Bengtsson and Eggum (1969), Just et al. (1983 b), Truelsen and Sorensen (1986) and Full- er et al. (1989).N fertilizer increases mainly the amount of storage proteins, hordeins, which are low in lysine, whereas the amount of albumins and globulins remains rather constant (Kirkman et al. 1982).Usually, however, the relative de- cline in lysine has been so slight that the total content of lysine in grain has increased (Jones et al. 1968, Thomke 1970).Arra clearly had the highest CP content of all the varieties studied.It has the highest CP content of the varieties com- monly grown in Finland and is mostly used as animal feed.Kustaa and Pokko, with their lower protein contents, are also used for distilling and malting.

In vitro digestibility
The variation in the in vitro digestibility of DM among barley samples indicated that there may have been differences in the energy value of the barley batches.DM digestibility was highest in variety Arra.Similarly, Just et al. (1983a) found differences in in vivo energy and organic matter digestibility between barley varieties.Thomke and Frölich (1968), however, found no differences in energy digestibility between high-and low-protein barley varieties.The energy digestibility and energy value ofbarley are usually neg- atively related to the fibre content (Bell et al. 1983,Jacyno 1995, Darrouch et al. 1996, Beames et al. 1996) but no such relationship was estab- lished in our experiment.We did, however, find a very narrow variation in the fibre content between our barley samples.
The high in vitro digestibility of N in Arra also indicated its higher feeding value.The high CP content of Arra explained some of the better in vitro digestibility of N, a positive relationship having been found between the protein content of the grain and in vivo digestibility in many oth- er studies, too (Eggum 1970, Eggum and Chris- Vol. 6 (1997): 295-303. tensen 1975, Taverner and Parrel 1981, Just et al. 1983b, Jacyno 1995).Fuller et al. (1989) showed that the apparent ileal digestibility of CP and most of the essential amino acids was higher in high-protein than in low-protein varieties of barley.In the experiment of Thomke and Frölich (1968), the differences in the feeding value of barley varieties were related to the CP content, a conclusion that is also in agreement with our results because Arra contained more CP than did Kustaa or Pokko.
The regression equation showed that CP, NDF and (3-glucan contents explained the varia- tion in in vitro digestibility ofN reasonably well.The relationship between CP and (3-glucan con- tents and protein digestibility was positive but NDF impaired protein digestibility.A similar relationship between protein digestibility and (3glucan content was found in the study of Bach Knudsen and Eggum (1984), because (3-glucans are mostly located in the cell walls of the en- dosperm, which also has the highest protein di- gestibility.Likewise in the experiments of Tav- erner and Farrell (1981) and Hall et al. (1987), hemicellulose and NDF, of which hemicellulose is the major component, were the most closely negatively related to protein digestibility and amino acid availability.

Pig performance
Barley variety had a greater effect on the performance of the pigs than did soil type or N fertilizer.The feeding value of variety Arra, measured by in vitro digestibility or pig performance, was slightly better than that of the other two varie- ties, Kustaa and Pokko.The diets composed of Arra contained slightly more lysine than did those composed of Kustaa or Pokko (5.5 vs. 5.3 g/kg).The daily lysine intake of the pigs on Arra diets was therefore higher than that of the pigs on Kustaa or Pokko diets.The combined effects of both increased lysine intake and higher digestibility resulted in the performance re- sponses of the pigs on Arra-based diets.The amount of soya bean meal in our diets was higher than that used elsewhere (Thomke andFrölich 1968, Thomke 1972) and may have masked some of the difference in the protein value of the barley samples.The DM intake and FCR of the pigs fed different barley varieties followed the same pattern as in vitro DM and N digestibility.The feed energy values of the bar- ley samples were calculated with the constant digestibility coefficients from feed tables (Tuori et al. 1995) and did not take into account the differences, if any, in the digestibility values.The actual differences in FCR calculated as feed units may therefore have been less than was reported.N fertilizer supply improved the FCR and carcass quality of the pigs in our experiment.Similarly, the higher N fertilizer supply of bar- ley has had a positive effect on the daily N re- tention (Just et al. 1983 a, b) and performance of pigs (Thomke 1976).Thomke (1970), however, concluded that the protein quality of barley de- clined with higher CP content since the relative content of most of the essential amino acids decreased in the protein.This finding is in agree- ment with our results, which suggest that extra N is mainly excreted in urine.Soil type did not have a marked effect on the nutritive value of barley.In other experiments, in contrast, soil type or growth locality has affected the chemical com- postion and nutritive value of barley (Thomke 1972, Just et al. 1983 a).In the study of Thomke (1972) growth locality affected both the content and digestibility of CP in barley samples.Simi- larly, Just et al. (1983a) found a difference in the digestibility of CP and energy between bar- ley samples grown on different soil types.How- ever, the differences were not very consistent because soil type and barley variety strongly interacted in both digestibilities.

Conclusions
Variety affected the feeding value ofbarley more than did soil type or level of N fertilizer.The in vitro digestibility of DM and N and the perform-ance of pigs were higher in variety Arra than in Kustaa or Pokko.N fertilizer increased the CP content of barley and slightly improved the FCR and carcass quality of the pigs.Soil type had only minor effects on the nutritive value of the barley studied.The in vitro digestibility of N was pos- itively related to the CP and P-glucan contents, and negatively to the NDF content of barley sam- ples.The results indicated that in vitro assay could be used as a preliminary tool to detect dif- ferences in the feeding value of barley samples.

Table 2 .
Effect of barley variety and sex on pig performance.LS means of factors are presented.

Table 3 .
Effect of soil type and nitrogen fertilizer level on pig performance.LS means of factors are presented.