Estimation of the protein value of pig feeds from fibre-bound crude protein fractions

Abstract

Crude Protein (CP) and Amino acids (AA) are essential nutrients for diverse processes in the pig's body therefore an adequate CP and AA supply for pigs is important to sustain performance and animal health. The utilisation of CP and AA depends on their digestibility and absorption. Therefore, methods to determine reliably the CP and AA supply to pigs are critical for precise feed evaluation. The evaluation of CP and AA supply has so far been based on <em>in vivo</em> determination of standardised precaecal digestible (pcd) CP (pcdCP) and AA (pcdAA) and <em>in vitro</em> estimates of pcdCP and pcdAA applying time-consuming and complex chemical-enzymatic methods. The objective of this study was to develop and establish a rapid laboratory method for estimating pcdCP and pcdAA based on the determination of CP (N · 6.25) insoluble in neutral-detergent (ND) or acid-detergent (AD) (NDICP, ADICP) and AA insoluble in ND or AD (NDIAA, ADIAA). <br/> A unique, large sample pool of 82 feed ingredients (cereal grains, differently heat-treated legume grains) was available on which <em>in vivo</em> pcdCP were determined in cannulated pigs. The CP was determined for cereal grains on the ND residue and for all protein feeds on the AD residue, because N compounds such as Maillard products or N bound to tannin or in phytate complexes are retained in the AD insoluble fraction. Crude protein was determined in feed ingredients and in their ND or AD residues. The concentrations of ND- and AD soluble CP (NDSCP, ADSCP) were calculated by difference of NDICP or ADICP to CP in feed. For the estimation of the concentrations of <em>in vivo</em> pcdCP for the entire dataset, a linear relationship was established between the concentrations of NDSCP or ADSCP and the <em>in vivo</em> pcdCP: y = 0.8640 (standard error [SE] 0.019) x - 13.37 (SE 7.479), where y represents the <em>in vivo</em> pcdCP (g/kg dry matter [DM]) and x represents the NDSCP (cereal grains) or ADSCP (protein feeds) value (g/kg DM). The coefficient of determination (R²) of this equation was 0.962. A validation with literature values showed a good fit of the equation to an independent data set (n = 20; R² = 0.955). <br/> For the estimation of pcdAA the same procedure which was previously applied to estimate pcdCP was used. Of the same sample pool 74 feed ingredients were available. Amino acids in feed ingredients and in ND or AD residues of feed ingredients were determined by an HPLC method. The concentrations (g/kg DM) of NDSAA, ADSAA were calculated by difference to total AA in feed. For the estimation of the concentrations of <em>in vivo</em> pcdAA for total AA and the entire dataset (n = 74), a linear relationship was established between the concentrations of NDSAA or ADSAA and the <em>in vivo</em> pcdAA: y = 0.823 (SE 0.018) x + 10.52 (SE 4.420), where y represents the <em>in vivo</em> pcdAA (g/kg DM) and x represents the NDSAA (cereal grains) or ADSAA (protein feeds) value (g/kg DM). The R² of this equation was 0.968. For the 17 individual AA the R² ranged from 0.895 to 0.984. This study shows that based on chemical analysis alone, namely determination of NDICP/ADICP and NDIAA/ADIAA, from which NDSCP/ADSCP and NDSAA/ADSAA are calculated, <em>in vivo</em> pcdCP and pcdAA values can be estimated with a standardised and rapid laboratory method.