A threshold for varroa infestation would be a good thing to have. It would define the moment for varroa treatments without any doubts, wouldn’t that be nice? You may already notice that I have some issues with this concept. But let’s start from the beginning.
There’s no doubt at all that treatments against varroa mites are important. I explained already why and how often. The aim is to keep colonies healthy and productive; there are some critical points. It’s important to know that the infestation doubles approximately every month, as long as there is brood. This relates to varroa biology, which is a story for another time. In this context, this means that during the productive season, varroa mites are happily increasing in number while the beekeeper can’t do anything about it. At least if he cares about the quality of his honey.
So, considering this doubling every month thing: we have to make sure that the colonies start into the productive season with an infestation as low as possible. On the other hand, we also have to make sure that the colonies make it through dearth periods without damage from the parasite. As most research on the topic comes from regions with temperate climate, these treatments are usually referred to as “summer treatment” or “winter treatment”. Alternatively as spring and fall treatment.
There are several diagnostic methods for varroa infestation. These were important in a time that not all colonies were infested with varroa. Nowadays, this mite is everywhere. Even the last stronghold against this parasite, Australia, fell last year. In Australia, it’s still good to do diagnostics to observe the spread of varroa mites and put adequate measures in place. But in most other regions of the world we can assume that there is varroa in the colonies.
The oldest method to check for varroa is to count the natural mortality of the mites. For that, you put a tray, sticky sheets or similar on the bottom board for some days and then count how many mites died in this period. This method is often referred to as “natural mite fall” – i.e. dying varroa mites.
Then, there are different “washing methods”. For this, you take an adult bee sample, and “wash” them with soapy water, alcohol, or icing sugar. The first two usually aren’t very popular as some bees die during the procedure.
Quality versus quantity
Knowing that varroa mites are there is one thing, this is a qualitative measure. The other is to know how many of them are in the colonies, i.e. having a quantitative measure. We need this if we want to establish a threshold for varroa infestation levels before treatment. In the beginning of the varroa era, there were some studies on this. When I began my PhD in 1994, the indication was that one mite on the tray per day meant 100-300 mites in the colony. Quite a range, isn’t it? Let’s say we count 5 varroa mites per day, this would mean from 500-1,500 mites in the colony.
Actually, this is the most accurate method we have. But it’s still not accurate enough. When you count varroa mites like this, then treat, and finally also count the mites after treatment, we have some correlation. But with an even larger range than that 100-300x.
The washing methods are much worse. If you take the effort to control the single bees for mites after washing them, you will still find quite a few of them. In addition, there’s the insecurity considering the relationship between the mites on the adults and those in the brood cells. This changes during the year. So, it’s impossible to say “If there are x mites on the bees, there are y in the colony”. Therefore, all diagnostic methods are qualitative, not quantitative, measures.
Why do we want a threshold for varroa infestation?
I think it’s important to ask why we even need to know the number of varroa mites in the colony. It’s time consuming, so it should serve a purpose, shouldn’t it? Usually, you get the answer “Because we want to treat before the infestation gets over a damage threshold!”. In principle, this is a good idea. It’s in line with what I always preach: When you see varroa mites on the bees it’s too late. And treat as much as needed and as little as possible. So, a threshold for maximum varroa infestation before anything bad happens would be great. In theory.
There’s just a little issue: we don’t know where this damage threshold is. Meaning how many varroa mites a colony can support without collapsing. And most probably it’s not the same in different locations and even between different years. It depends on factors like:
- How long is the breeding season of the colony in a certain area (i.e. how long can varroa mites reproduce),
- how much brood do colonies have over the season (i.e. the opportunities for the parasite to reproduce),
- how many colonies are there in the area (i.e. the risk of reinfestation).
- How is the nutritional state of the colony? Well-nourished colonies with good pollen supply support higher infestations. Less obvious but not less important.
- What is the viral load in the colonies? Because DWV and other viruses are associated with varroa. There might be a low varroa but a high viral load. Which makes colonies collapse.
What do the data tell us?
I’m very much in favour on deciding based on data. But we also have to ask ourselves if the data we have answer the questions we need answers for. We know that colonies with a certain infestation have a higher risk of dying over winter (or another dearth period). But even this moves within a range. I know numbers for Germany, they will look different all over the country and even more in other regions.
Collecting data only makes sense if you’re able to use them. Considering everything I told you above, I think it’s useless in this case. There are two other points I want to make in this context:
When the idea of a damage threshold came up, the associated virus infections didn’t play a large role as they do today. Today, we know that varroa is a catalyst for DWV infections and that this and other viruses are often the straw that breaks the camels back. Considering the high insecurity in predicting varroa numbers from the diagnostic methods we have, estimating the viral load is nearly impossible.
Thresholds for varroa infestation give a deceptive sense of security
In addition to that, we need to treat BEFORE we reach a threshold of any kind. I.e., before the varroa and virus complex damages the colony. In my opinion, thresholds for varroa infestation give a deceptive sense of security. As I explained above, what number of varroa mites damages the colony depends on quite a few factors. I’m also pretty sure that there are some I don’t even think about. Biological systems – like the honey bee-varroa-virus complex – aren’t binary. There isn’t a magical number that fits all situations.
To be honest, I formed this opinion quite recently. Mainly because I talked to beekeepers from many different regions. Which don’t always have temperate climate. As mentioned above, the treatment schemes were developed in those regions. Nothing bad about that, but obviously in subtropical or Mediterranean conditions things look differently.
Teaching, I often heard from beekeepers: “We don’t have broodfree winters, but may have a period without brood in summer!”. Or “We have brood all year long and several population peaks during the year!”.
Colony development depends on climatic conditions
These conversations always made me think about a generalized recommendation, that doesn’t rely so much on “summer” or “winter”. There are always things that are generally valid, like the biology of honey bees or varroa mites. And others that change. Like the development of the colonies, which depends also on the climatic conditions. Let’s take a look on the “baseline”, temperate conditions.
Temperate conditions as a “baseline”
In this case, we have four defined seasons:
- a cold winter without brood and food resources,
- a warming spring with plenty of food in which the colonies grow and reach their population peak at the end of it,
- a warm summer, still with precipitations, the colony strength begins to decline, especially the amount of brood. Food resources are still present, though not as much as in spring.
- a mild autumn, with increasing precipitation and cooler temperatures. Food resources begin to be scarce, the colonies reduce brood but are rearing longer-lived winter bees.
In these conditions, the summer treatment makes sure that the winter bees grow up with a very low varroa infestation and viral load. The winter treatment prepares the colonies for the productive season – when that doubling every month occurs.
“Deviations” from this baseline
In Mediterranean or subtropical conditions things look different. In Sicily, Lebanon, or California we have a mild and rainy winter, food explosion in spring, and a hot and dry summer. There’s brood in winter, but colonies may stop in summer. There’s a strong development in spring, with a population peak and maybe a second one in autumn when the resources come back. Not as big, though. In subtropical conditions like in Florida, there’s brood all year round. There are population and productivity peaks, but not as marked as in the former cases. There are still dearth periods, though, like a “dry season” that can vary in length and severity.
Common features rule the treatment schedules
This shows very briefly what all conditions have in common and what is different. So, what are the common features that relate to varroa treatments? Two things: population peaks and periods of dearth. Colonies are most productive during the population peaks and most vulnerable during the dearth periods. This means that:
- We need a treatment to keep the varroa number low during the productive season. Ideally, this happens in a period with no brood, i.e. before colonies begin to grow. You can use natural brood interruptions or cage the queen to induce it. The brood after this break will grow up healthy – and be productive foragers as adults!
- The second treatment must prepare the colonies for the vulnerable period, i.e. when the main honey harvest is over and the colonies begin to decline in strength. This would correspond to the “summer treatment”.
This is adaptable to different conditions and relies on what every beekeeper deserving that name does: Observing the colony development. Takes less time than counting mites and is more flexible and reliable than that and a non-existing damage threshold for varroa infestation.
As I’ve been talking about this quite a bit in this period, I’m noticing that I’m not totally alone with these thoughts. Others contradict vehemently. But, there are data that support this concept. Together with some colleagues, we’re working on strengthening the model. Stay tuned.
This post gives the background to a – controversial LinkedIn post I made. Follow me there if you want to be part of the discussion. If you want more background knowledge on bee health, check out my Bee Health Compendium!