Re: Hive postmortem; NYC rooftop [pics]
The problem with diagnosing parasitic mite syndrome is that the mites which do all the damage are long gone by spring when you find a dead hive.
The typical course of events in a healthy hive is for them to build up in the fall, cluster without brood until the days start getting longer and it starts to warm up in late January (at least here in southern Indiana, dates will vary with lattitute and altitude). Once spring is on the way, the queen will start laying a fairly small batch of brood, and really take off once pollen and nectar start coming in, which usually means late Feb/early March here.
In a hive heavily parasitized by varroa mites, the bees are infected with several female laying mites per cell in the fall when winter bees are being raised in somewhat smaller numbers than typical summer brood, with the result that they are often deformed by both being fed on by too many mites and sickened or deformed by the various virus diseases the mites are vectors for. The emerging winter bees are neither as well feed nor as well developed as un-parastitized bees, with the result that they die off early in winter instead of living until late spring, and are also protein deficient. This combination causes excessive losses of bees during the brood break -- the mites tend to die off then, so you don't see them any more -- and then later in winter when brood laying commences, the cluster is too small to support enough bees. Protein deficient bees starve trying to feed the brood, the cluster is too small and shrinks more, so they die of cold along with the brood.
If there are still significant numbers of phoretic mites in the hive, the emerging brood will be deformed, but the real key to diagnosis is the presence of mite droppings in the form of small white specks in the brood nest. When the cluster is too small, they cannot clean all the cells properly and the white specks persist after the bees are gone.
There is nothing you can do to save the hive after September, because the damage has been done by then. You must do a complete mite count (sugar roll, ether roll, maybe powdered sugar drop) to see, and if you see more than a couple mites, treat. You must have very low mite counts when winter bees are developing or they won't make it through the winter.
Inadequate pollen reserves will produce the same results, but you won't have mite droppings in the brood cells in that case, and often have low bee population in the fall.
The strain of bees doesn't have much to do with this problem -- Russians are supposed to be much more mite resistant, but will still die from protein deficiency or a large mite population if they don't manage the mites well.
The OP didn't have a pollen deficiency, so it's much more likely the immediate cause was mites. Could also be a weak queen or bad genetics, hard to tell.