13 sources, every one opened at its publisher — and six of them are open access, so you can read the paper itself for nothing and check this book against it. A varroa mite does not drink your bees' blood. It eats the organ that lets a winter bee live six months — and it eats it inside a sealed cell, before she has emerged. For fifty years the mite was understood to drink haemolymph, the way a tick drinks blood. It does not. It externally digests and consumes fat body — the tissue the researchers who found this describe as “roughly analogous to the mammalian liver”. In a bee that one organ stores the protein she overwinters on, runs her immune response, detoxifies pesticides, and decides whether she is going to be a short-lived bee or a long-lived one. In the feeding trial that settled it, mites fed haemolymph showed fitness no different from the starved control. Mites fed fat body lived longer and laid more eggs. Why that changes which week you act in Winter bees are not summer bees living quietly. They are provisioned to last five or six months, and the provision is held in the fat body. They are reared in a specific few weeks at the end of the season — which is exactly when the mite population is at its height. A study published in 2004 measured the result: bees infested during the pupal stage “do not fully develop physiological features typical of long-lived wintering bees”. The damage is done before the bee emerges. And the cost shows up in days. Following 29 colonies and collecting the dead daily, winter bees in the colonies that failed lived a mean of 71.3 days against 88.4 in the colonies that survived. The two groups were indistinguishable for the first fifty days. The colony that dies in February looks fine in October. The finding this book is named for National survey data from England and Wales — 4,339 beekeepers, 37 counties, 18,700 colonies — tested treatment timing against treatment product. The result: “Lower overwinter colony losses are predicted by correct treatment temporality, but not by product application alone.” Applying a treatment, by itself, did not predict lower losses. Applying it at the right point in the mite's cycle did. And over one third of beekeepers were found to be missing those windows. From the other side of the Atlantic, in the worst United States loss year on record: losses did not significantly differ between users and non-users of amitraz. If the story were simply that the usual product had stopped working, the people who had stopped using it should have done better. They did not. And the explanation everybody gives that the data does not support Ask any beekeeper where the autumn mite surge comes from and you will hear about mite bombs — a collapsing colony nearby whose mite-laden bees drift into yours — or about robbing. Both were tested by tracking bees between 32 colonies across eight apiaries. Mite growth was associated with visitation by non-natal bees in general (P = 0.03) and not with bees from the high-mite colonies specifically (P = 0.19). More bees arrived from the low-mite donors than the high-mite ones. The authors state plainly that the results “do not support the current two leading theories regarding mite immigration”. One intervention did measure: colonies without robbing screens grew mites faster than screened neighbours (P = 0.01). What is in the 52 pages A 14-row failure finder, arranged by what you saw rather than by what was wrong — find the row, read the chapter. 10 full-page plates, including the mite in the position it was actually measured in: wedged under an abdominal plate on the underside, not riding on the thorax where almost every photograph shows it. The denominator trap — why your mite percentage can double with no mite added, because the bottom of the fraction collapses every autumn. Reinvasion, in numbers: 266 to 1,171 mites walked into each colony over three and a half months where more than 300 colonies stood within flight range, against 72 to 248 where neighbours were few. Resistance, properly understood — 90% of apiaries in one survey, but from 95.7% to 31.8% within a single province, and the French mutation absent entirely. Plus why a recessive trait in an inbreeding parasite fails suddenly rather than gradually. Thirty years, 274 efficacy values, 32 countries, and no robust evidence of resistance to the one compound that is not a lock a mite can pick. The sealed-cell timetable, which is the arithmetic every treatment decision actually rests on. Three printable worksheets: the count log, the treatment record, and your apiary's neighbours. Every recommendation labelled Measured, Mechanism or Practice, so you can see which is which at a glance. What it does not do It does not replace the free guide, and it says so on page one. The Honey Bee Health Coalition's Tools for Varroa Management is free, revised regularly and genuinely good, and it is the reference for sampling protocol and for what may be applied. This book cites it rather than repeating it. What a management manual cannot do is explain why a beekeeper who followed it still lost the colony. It names no product as approved. Which varroacides are permitted differs by country and changes. This book explains classes of treatment and what each one can physically reach given the brood cycle; your own national register says what you may use. It is not veterinary or legal advice. And it has a chapter on what it does not know. Three things, named: the sugar-roll recovery percentage everybody quotes and nobody sources; a famous pair of figures in a paper that could not be opened at the publisher; and resistance outside the regions actually surveyed. It also reprints, twice, the one finding that cuts against its own argument. What you get PDF — 52 pages, US Letter, prints cleanly at home. There is deliberately no EPUB: the value is full-page plates, a finder table and printable worksheets, and a reflowable format destroys all three. 10 plates, 9 figures drawn from the published numbers, 14 failure modes, 3 worksheets and 13 numbered sources with a note at the foot of every page that uses one. Instant download. Nothing ships. The ten plates are digitally rendered illustrations of the states described. They are there because varroa work is done by eye, and the book says so in its colophon.