Amino acid and energy digestibility of Brassica napus canola meal from different crushing plants fed to ileal-cannulated grower pigs

Canola meal (CM) can replace soybean meal (SBM) as dietary source of supplemental protein for pigs; however, varying nutritive quality of CM may cause inaccurate diet formulation and unexpected growth performance, but is rarely reported. In an 8 × 8 Latin square, the standardised ileal digestibility coefficient (CSID) of crude protein (CP) and amino acids (AA), apparent total tract digestibility coefficient (CATTD) of gross energy (GE), digestible energy (DE) value and the predicted net energy (NE) value were evaluated for 5 Brassica napus CM samples originating in a single crop year from 5 Western Canadian crushing plants and compared to SBM. Eight ileal-cannulated grower pigs (initial body weight [BW]: 32 ± 1.9 kg) were fed 8 diets (400 g SBM/kg, five 400 g CM/kg, wheat-based basal and N-free) at 2.8 × maintenance (0.46 MJ of DE per kg of BW0.75) for 8 periods of 9 days each. Standardised to 100 g moisture/kg, the SBM and 5 CM samples contained 461 and 357–413 g CP/kg, 23.9 and 10.6–35.8 g ether extract (EE), 69 and 218–251 g neutral detergent fibre (NDF)/kg and 17.6 and 17.9–18.1 MJ GE/kg, respectively. The CSID of essential AA such as lysine, threonine and methionine was greater (P < 0.05) in SBM than in the 5 CM samples, except for tryptophan. The CSID of lysine, threonine and methionine differed (P < 0.05) among CM samples and ranged 0.84–0.88 for lysine, 0.82–0.86 for threonine and 0.89–0.92 for methionine. The CATTD of energy ranged 0.753–0.788 and predicted NE value ranged 7.56–7.97 MJ/kg as fed among CM samples. Total glucosinolate content ranged from 1.2–7.6 μmol/g and was negatively correlated (r = −0.390; P = 0.013) with DE value of the 5 CM samples, but not with CSID of AA. In conclusion, samples of CM processed in Western Canada varied in CP and EE content, but CSID of AA varied minimally. The measured DE and predicted NE values were close to NRC (2012), but greater than values for rapeseed meal in the INRA database (Sauvant et al., 2004), indicating that database selection for diet formulation matters.