Resumen

PINOS-RODRIGUEZ, Juan Manuel; GONZALEZ MUNOZ, Sergio Segundo; MENDOZA MARTINEZ, Germán David  y  MARTINEZ GARZA, Ángel. Análisis estadístico de experimentos de digestibilidad in vitro con forrajes. INCI [online]. 2002, vol.27, n.3, pp.143-146. ISSN 0378-1844.

Four statistical designs were used to analyze data from an in vitro DM degradation of alfalfa and rye-grass hay using the first phase technique of Tilley and Terry at 48h, duplicated (two runs) and three tubes (repetitions) for each forage (six tubes for each forage). The following models were used: 1) means from each incubation, and two repetitions for each forage; 2) each tube was a repetition, with six repetitions for each forage; 3) same as model 2 and the term incubation in the model, which was not used as an error term; 4) each tube was a subsample and incubation x forage was an error term in the model. Forage digestibility was different (P<0.05) with the first three models; model 1 has only two degrees of freedom to estimate error, whereas for model 2 and 3 there are 10 and 8 degrees of freedom, respectively. However, models 2 and 3 do not take into account the variability of experimental conditions which are found on in vitro digestibility trials and, as a consequence, there is a higher probability of obtaining non-repeatable results. Analysis of variance done with model 4, which used interaction incubation x forage as an error term, did not show differences (P>0.05) for the forage. Therefore, if the experiment is repeated, there is a lower probability that the results can be repeated also. These results suggest that model 4 would be more appropriate to analyze in vitro digestibility trials because it takes into account the variability of experimental conditions which are found on in vitro digestibility and determines the repeatability of the experiment.

Palabras clave : Análisis Estadístico ; Digestibilidad in vitro ; Forrajes .