Abstract
1. This holo-analysis of the effects in turkey nutrition of Bio-Mos® (BM) (Alltech Inc., Nicholasville, KY) aims to elaborate comprehensive empirical models for the assessment of responses under practical conditions and for its comparison with other potential pronutrient antibiotic replacements.
2. The data bank utilised was obtained from 54 publications on BM in turkeys from a total collection of 57 papers (1995–2003) on the use of saccharide products as turkey feed additives. These yielded 33 start-to-finish negatively-controlled feed intake, liveweight gain and feed conversion and 24 mortality effects in the USA, France and Poland, using 12 723 birds with a mean of 212 birds per treatment.
3. The feed, gain, conversion and mortality responses to BM compared with the relevant negative controls average −243 g, 57·0 g, −0·0156 and 1·29% with coefficients of variation of 451, 451, 705 and 426%, respectively. Beneficial gain, conversion and mortality effects were found in 52, 52 (33% jointly) and 52% of the tests, respectively.
4. Holo-analytical multiple regression models comparing conventional P ≤ 0·05 in/P ≥ 0·10 out and less stringent P ≤ 0·25 in/P ≥ 0·34 out partial regression coefficient standards, with and without ≥1·5 × root mean square error outliers, contain significant independent variables for negative control performance, duration, year of test, not-USA test, discontinuous dosage, cage housing, mash feed, feed antibiotic, feed anticoccidial, disease condition and high control (>10%) mortality.
5. These developmental models indicate fruitful pathways for future developments with an accent on the definition of specific dose–response relationships, whilst also confirming the value and future potential for BM in turkey production in praxis.