>I also recall reading a study where higher mite mortality was attributed to mites being smothered in tighter fitting brood cells, but, I can't recall the source.
http://www.apidologie.org/articles/apido/pdf/2002/01/Martin.pdf
The conclusions:
"For ectoparasities which reproduce in enclosed cavities the amount of space can be an important constraint on their ability to reproduce successfully. Therefore, species like Dichrocheles phalaenodectes which breeds within the tympanic organ of moths (Treat, 1975) and Varroa sp., display traits such as lack of cannibalism, nest sanitation and space partitioning (Donzé and Guerin, 1997).One consequence of space partitioning in Varroa sp. is that the first (male) egg is laid near the cell cap. This increases the survival probability of themalemite since it is the only place in the cell not affected by the bee’s molt (Fig. 2). However, the male mite must now pass the constriction caused by the bee’s appendages to reach the feeding site which is established by the mother mite on the bee’s abdomen (Fig. 2). Since only one male is produced per batch of eggs, its death will result in all the female offspring being unmated and so unable to produce offspring (Akimov andYastrebtsov, 1984; Donzé et al., 1996; Martin et al., 1997; Harris and Harbo, 1999).
"A survey of the literature revealed a close correlation (r2 = 0.97) between fore wing length and brood cell diameter across 14 races of A. mellifera (Fig. 1), also fore wing length is closely correlated to bee head width (r2 = 0.97 worker & drone) in Apis (calculated from data inRuttner, 1988). Therefore, since the pseudo-clone which is among one of the larger A. mellifera races, is being reared in some of the smallest cells found in A. mellifera. (Fig. 1), there will be significantly less space between the bee pupae and cell wall in cells occupied by pseudo-clones than A. m. scutellata workers which may impede the movement of the mites. This may explain our frequent observations that dead male protonymphs and some dead mother mites appeared to be trapped in the upper part of cells containing the pseudo-clone. This is illustrated by the high level of male protonymph mortality found in cells occupied by the pseudoclone (48 × 0.90 = 43%) compared to those occupied by A. m. scutellata workers (28 × 0.59 = 16.5%). While in A. cerana drone cells, ancestral host of Varroidae, only 1–2% of the male offspring die (Tab. II). Interestedly this species builds the widest drone cells (7.1–7.2mm)of any Apis sp. but rears the smallest Apis drones based on head width.” --Reproduction of Varroa destructor in South African honey bees: does cell space influence Varroa male survivorship? Martin, S.J."
Followed by this paragraph which is only assumptions not based on anything observed in the study other that the fact that they WERE affected by the space:
"Although reproduction of Varroa sp. is affected by the space between the developing bee and cell wall, reducing cell sizes as a mite control method will probably fail to be effective since the bees are likely to respond by rearing correspondingly smaller bees which explains the close correlation between cell and bee size (Fig. 1)." --Reproduction of Varroa destructor in South African honey bees: does cell space influence Varroa male survivorship? Martin, S.J."