Detecting genetic differences between groups of sheep during parasite infection.

When experiments are established to compare parasitological differences between genetic groups of animals, sampling of families within groups and individuals within families are both important. Many parasitological measurements are quite variable, so relatively large numbers of animals are required to detect small differences. Calculations are presented which show, for example, that 100 animals per genetic group will only have an 80% chance of detecting a difference in faecal worm egg counts as significant (P < 0.05) if the real difference between the groups is about 3-fold. When sampling genetic groups such as flocks or breeds, it is preferable to represent as many families as possible.