The following parameters should be taken into account:
1. Do not include weak colonies or colonies affected by diseases other than varroa in the study.
2. Equalize or randomize bee breed depending on the aim of the test regarding genetic diversity.
3. Select sister queens or unrelated queens of same age.
4. Measure and equalize colony strength (see the BEEBOOK paper on estimating colony strength (Delaplane et al., 2013)).
5. Measure and equalize the amount of brood (see the BEEBOOK paper on estimating colony strength (Delaplane et al., 2013)).
Presence and type of brood is determined by the mode of action of the product. The tests should thus be performed in the absence of sealed brood, unless the product is intended to be effective on mites in capped cells.
6. Initial varroa infestation level should be high enough (> 300 mites per colony) to be able to measure mite drop.
It should however be below damage thresholds (e.g. for central Europe: <3,000 mites per colony, see also Table 5 and section 4.4. ‘Estimating damage thresholds’) and comparable between hives included in the study.
7. Treatment history should be similar for all colonies to equalize the effect of past treatments on the results (e.g. type and amount of acaricide residues present in the wax); when possible use residue free wax.