Plates were spread in triplicate and the direct counts averaged to estimate the total heterotrophic bacterial population.
Three samples of each plasmid were tested in triplicate.
This counting was done in triplicate, using three independent samples of the same population.
The same serum samples were tested in triplicate to confirm reproducibility of the results.
Faecal cultures were set up in triplicate (1995-1996) or quintuplicate (1997) at the beginning of each series of the plot trials.
Each bar represents the mean of three independent determinations performed in triplicate.
Data are the means of experiments performed in triplicate.
Specimens for identification were collected in triplicate from all species in the plots.
