I agree with burns375, I think more or less contant sampling is exponentially more effective at monitoring than any one-off per season testing, of any kind.
That's one of the reasons I am so sold on sticky boards -- because sampling using that method causes no disruption to the bees' lives at all, and very little effort on the part of beekeepers, either. All it takes is a beeks' determination to do it. And if you set up your equipment correctly, pulling the boards can even be done in the dark when you get home from work.
And since it takes no skills beyond spreading cooking oil on a sheet of plastic and then picking up the mites that fall and arranging them in groups of five for easy counting, anyone can do it. And I think this is particularly important for new beekeepers who are just getting used to handling bees. Digging into your hives enough to collect the right bees for doing any of the roll-type testing, requires suiting up, lighting your smoker, smushing some bees, etc. The result is that it isn't done as often as needed, or even at all, like it ought to be. So vital information is not known until a dead-out makes the point irrefutable.
In answer to the OPs question: Last summer I used the numbers published by NY BeeWellness and the similar Canadian sources, since I am in northern NY. IIRC, NYBeeWellness had a late summer/fall need-to-treat threshold of 20 mites/day and the Canadian one uses 12 mites/ per day. My counts were in the single digits all summer until (in the first week of Sept.) one of my hives produced 15 mites/day, repeated over several days. I treated that one. About ten days or so later another one spiked so I started to treat that one, too. My third hive stayed below the threshholds so I did not treat it. The first hive treated got a full dose of Apiguard, but the latter one only got a half-period treatment because temps got too low to continue before it was done.
I have continued to monitor since then, including all winter even though there are no published mite/day thresholds for the broodless period of a deep northern winter. (For curious minds, the fully treated hive's long term winter average is less than 3 mites per day, the half-treated one's average is less than 1 mite/day and the untreated one's average is 0.3 mites per day.) In the winter some of my test periods have been 30 days long because it is a such pain to unwrap the insulation layers to access the sticky board slots. But whenever I have the panels off, I pull and count. Next year's insulation arrangements will allow regular monitoring, because I think it's interesting, and potentially useful. In the next month when I will start to remove some of the outer layers of my extreme-climate insulation, I will once again start regular, even daily counts.
Oh, and, one other thing about sticky boards, aside from actually doing the counts, the other critical thing is: You have to record your numbers. Because you won't remember what they were. Most of the value of the information you are collecting is the change over time, not the actual number, or treatment threshhold. I'm terrible about recording data, so this is the hardest part for me.
Enj.
