A comparison of calculated and determined apparent ileal digestibility of amino acids in cereal-soyabean diets for young pigs

The digestibility experiment was carried out on twelve cannulated pigs between 14 and 27 kg body weight. The animals were fitted with a PVTC cannula and fed with two starter diets, A and B, differing in metabolizable energy content (13.5 and 14.5 MJ ME/kg, respectively), with no supplemented amino acids (AA), in order to measure digestibility coefficients of protein-bound AA. The diets were formulated to contain 0.6 g ileal digestible lysine/MJ ME and were calculated on the basis of AA analysis of ingredients and their tabular values of apparent ileal AA digestibilities. For both diets, differences between the calculated and determined digestibilities of the limiting essential AA were found. The determined digestibilities tended to be lower than the calculated ones. Of the essen­ tial AA, the differences were greatest for threonine (6.8 and 9.5% of calculated values for diet A and B, respectively) and lysine (4.3 and 8.3% for diet A and B, respectively); the differences for methionine and tryptophan ranged from 4.0 to 6.2%. In conclusion, the present study shows that the digestible A A supply in a diet can be predicted from total AA content determined in the individual ingredients and from average literaturę values recommended for their AA digestibility, however, it is necessary to accept an 8-10% error for calculated values.


INTRODUCTION
The nutritive vałue of protein in feedstuffs is not only determined by its total amino acid (AA) content, but also by AA digestibility, with particular reference to the limiting AA.For formulation of pig diets, apparent ileal AA digestibility coef-ficients of feeds, presented as average values of many literaturę data, are recom mended (e.g., Rademacher et al., 1999).However, considerable variations in apparent digestibility values of AA in many feeds were reported.For example, lysine digestibility in wheat samples ranged from 59 to 80% (Sauer and Ozimek, 1986;Fan et al., 2001).Large variability was also found for other feeds like barley (Sauer and Ozimek, 1986) rapeseed meal (Buraczewska et al., 1987) and soyabean meal (Fan et al., 1996).The aim of our studies was to formulate diets for piglets containing a determined level of apparent ileal digestible AA, e.g., 0.6 g digestible lysine/MJ ME.For this purpose, the chosen cereal and high-protein ingredients were analyzed for their total AA content and their levels of ileal digestible AA were calculated using the digestibility coefficients recommended by Degussa-Hiils (Rademacher et al., 1999).The objective of this study was to compare the calcula ted and determined content of ileal digestible AA in order to test the adeąuacy of tabular coefficients for diet formulation.Ileal digestibility of dietary AA was determined on cannulated pigs.

Animals and experimental design
The experiment was carried out on twelve castrated małe crossbred piglets (synthetic line 990), surgically fitted with post-valvular T-caecum cannula (PVTC) according to van Leeuwen et al. (1991).After cannulation the piglets were fed two experimental diets (6 pigs per diet) during two experimental periods at about 15 and 25 kg body weight (BW), according to the change-over design.Each period consisted of at least 7 days adjustment to the diets, followed by 3 days (12 h each) collection of ileal digesta.

Formulation of diets and feeding program
Two starter diets A and B, containing 13.5 (A) and 14.5 (B) MJ ME/kg and 0.6 g ileal digestible lysine/MJ ME were formulated on the basis of chemical analysis of dietary ingredients, including AA (Table 1), their tabular ileal digesti bility coefficients (Table 2) and "ideał protein", recommended by Degussa-Hiils for pigs (Rademacher et al., 1999).Apart from digestible lysine, all other essential AA were balanced in the diets composed of seven protein-containing ingredients, as described in Urynek and Buraczewska (2001).Chromie oxide (Cr 2 0 3 ) was used as a marker.The pigs were fed two equal portions daily at 08.00 and 20.00 h in accordance with BW (5%).The unpelleted feed was mixed with water (1:1) just before feeding.Water was supplied ad libitum.

Sampling, analysis and calculations
Ileal digesta was collected using bags attached to the cannulas during the three days for 12 h per day, between 8.00 and 20.00 h.The bags were changed approximately every hour and their contents were immediately frozen in plastic containers at -20°C.After each collection, samples were thawed, pooled per animal within each experimental period, freeze-dried and ground (ty 0.5 mm) before chemical analysis.Cr 2 0 3 in samples of feeds and freeze-dried digesta was analyzed by the method of Fenton and Fenton (1979).Dry matter, nitrogen, ether extract, crude fibrę, total starch and ash were analyzed using standard methods (AOAC, 1990).The content of NDF and ADF was determined using a Fibertec System (according to the instruction).
AA were analyzed with a high-pressure amino acid analyzer, Biochrom 20, according to Llames and Fontaine (1994).Methionine and cystine were deter mined after oxidation with performic acid.Tryptophan was determined by reversedphase-high-performance liquid chromatography (HPLC) following alkaline hydrolysis with barium hydroxide.The apparent ileal AA digestibilities were deter mined according to the relevant eąuations (Rademacher et al., 1999).ME of die tary ingredients was calculated from the determined chemical composition of feeds (Table 1) using a corrected eąuation (Hoffmann and Schiemann, 1980).

RESULTS AND DISCUSSION
Determined AA digestibility values were similar for diets A and B (Table 3).There were differences between the calculated and determined digestibilities of the limiting essential AA in both diets (Table 4).The determined digestibilities tended to be lower than the calculated ones and these differences were more pronounced in diet B than A. Of the essential AA, the differences were largest for threonine (6.8 and 9.5% for diets A and B, respectively) and lysine (4.3 and 8.3% for diets A and B, respectively).For both diets, smaller differences ranging from 4.0 to 6.2% were observed for methionine and tryptophan.It is commonly known that many factors, including methodological approaches, fineness of grinding, dietary AA level and AA distribution among protein fractions (e.g., in cereals), NDF content (affected by genotype and growing conditions), temperaturę used in feed processing, level of antinutritional factors, and content of undigested endogenous protein at the ileum, are likely responsible for the variation of AA digestibility values reported in the literaturę.Additionally, there may be associative effects that cause the digestible supply of AA in a mixture of feeds to differ from the sum based on the digestibilities determined in the individual ingredients.Imbeah et al. (1988) reported that there were no differences between the calculated and observed digestibilities of the essential AA in a feed mixture of barley and soyabean meal, however, there were significant differences in lysine (10.5%) and phenylalanine (5.4%) content in a mixture of barley and rapeseed meal.
In conclusion, the present study shows that the digestible AA supply in a diet can be predicted from the total AA content determined in the individual ingredi ents and from average literaturę values recommended for the particular AA digest ibility, however, it is necessary to accept an 8-10% error of the calculated values.

TABLE 4
Calculated and determined content of digested (apparent, ileal) amino acids in diets A and B, % as-fed