We all have seen the effects of mites on honey bees, and many beekeepers are in real trouble, having lost all or most of their bees over the last few years. Many of us have tried various ‘cures’ to control or reduce mite numbers but much of the information is based on opinion of hearsay and little, if any scientific evidence. And what is science, anyway? Webster’s defines science as a “systematized knowledge derived from observation and experiment.”
So, let’s apply some scientific principles when looking for that magic bullet for mites. If you are going to test some of your own ‘cures’ this year, here are some guidelines on the correct procedures for conducting experiments and making reliable observations. If you consider them carefully and follow the steps, you will not only have done some good scientific research, but will begin to understand how time-consuming and expensive this work is and why it takes so long for a mite or any honey bee research to produce reliable results that can be published.
1. Forming a hypothesis
The first step in forming a hypothesis is to ask a question. An example of this is: Do basil leaves kill mites? After doing some experiments you will come up with some answers or potential explanations (hypotheses). But this simple question has many complex layers, such as: Do basil leaves kill mites, bees, brood? Do they have an effect on honey production? And which mite does it kill? And which bees are we talking about?
So you have to ask a very specific, precise question, taking into account all these other layers. A more specific question then, taking into account all these other layers. A more specific question then, and the goal of your research would be: Do fresh basil leaves control Varroa mites in Carniolian bees in my state this year? Now you’re ready to fo the next step.
2. Designing an experiment
How do you set up an experiment to answer you question? First, you have to go through a check list to define the parameters of your test.
You must include in your experiment colonies that do not receive any treatment (fresh basil leaves) at all. These are called control colonies. They get everything else (Terra, Fumidil-B, etc.), except the leaves. This is very important for without comparing tour tested-colony results to no-treatement control colony results, your experiment is meaningless. Further, make sure everything else in all colonies is equal: All should be of equal strength, a 3-lb. package, all have the same type and aged queen (3 months, old, Carniolan, same source) and all are treated equally (fed the same, worked the same day, medicated the same amount); the only difference is the one thing you are looking at – basil leaves. Then you need to decide how many colonies are needed to make your results meaningful.
It won’t be very impressive if you use only two colonies (one treated, one not), because any results you get could simply be due to chance. One hundred colonies would be best, but most of us are not able to do this, so if possible compromist on 10 or 20 colonies per treatment (are you beginning to see why research costs so much?). That means 10 colonies will receive the basil and 10 colonies will not. If you want to test two different amounts of basil, that will require an additional 10 colonies in the test. These colonies should not be used for any other purpose until the experiment is done.
B. Test Product
What is the material you are testing, how do you apply, how much and how often do you treat? In the case of our basil leaves, you could come up with fresh, chopped basil leaves, scattered on the top bars of the broodnest, in the cup an d half cup quantities, once a month. (Weight is better than volume, if you can.) It is better to have at least two different amounts of material, to give you more options. If you just test basil leaves and no basil leaves, and both treatments die you h ave not answered your question. But if you test two (or more) amounts of basil applied to colonies you will get more information.
This is how many pesticides are tested – not if it just kills insects, but how much of it is needed to kill all of the inse cts, some of the insects, none of the insects and not kill the plants they are on. The reasons you absolutely must carefully plan, set up and carry out your project, and take very good notes, plans, and results should be able to duplicate your experimen t to see if it works for them. The sure sign of good research is ‘repeatability’ by anyone (nearly) anywhere. Now that you have decided on what tou are testing, and how you are going to do it, you can go to the next step.
3. Collecting Data
Okay, you think you are ready to start? Nope, you have to make some more decisions. One is how often will you make your observations? This means collecting bees and counting mites, or, in research-speak, collecting data. Data are merely the numbers of mites counted per colony, per treatment, per month. Since you cannot count all of the mites or all of the bees in each colony, everytime you collect data, you have to take a representative sample. So, you need to decide what a sma ll will be. How large will a sample be or within what range should it be? For instance, if you are collecting bees, a sample could be no fewer than 50 bees and no larger than 200 bees.
Next, how often will you collect your samples? Many Varroa researc hers sample bees once a month durring the summer and into the fall months, while others test once a week; you have to decide for yourself how often and how long you can realistically sample you colonies. Remember, to get good results you can’t skip data collection times. Now, cancel those vacation plans before the airline penalizes you!
Next decide on how you will collect those samples. If you are looking for mites on bees, will you use the ether roll method, tobacco smoke and sticky board, or collec t adult bees. You don’t want to use an Apistan strip to collect mites because even a short exposure could affect the mite population such that a basil-leaf test would be meaningless. This will depend on how much time you can spend collecting and what yo u will collect in it. Once you choose a sampling technique however, use it throughout the entire experimental period on every colony the same way every time.
Remember, make all other conditions as identical as possible, collect data on the same day, in the same way and treat all colonies alike except for that one difference, the treatment. For instance, if you have to treat some colonies for Nosema, treat all colonies with Fumidil-B, since you are not doing Nosema research, but Varroa mite research. Wh en collecting the data, you will need to devise some kind of data collection sheet to keep sampling dates and treated colonies in order.
Next, pick your apiary site. There can be much variation even in defferent apiary locations. So, to keep it simple, use only one apiary for your experiment if at all possible. If you must use more than a single location be certain you have colonies in each spot receiving the same treatment. You’ll need the same number of treated colonies and controls in each locatio n. It’s best to keep everything together in one site though. Once your location is picked, your colonies equalized, your levels of basil leaves chosen. (full cup, half cup, control) and randomized, take your baseline data. Establishing a baseline simpl y means “what is the condition of each colony before the treatments began.
Take samples, as described above, before you start your treatment. In our experiments we started with three-lb. packages. For our purposes let’s assume no Apistan treatments wer e given at installation. Then, our experiment will begin some time (pick a date) in June. Since probably will not be able to count mites until later in the year, devise a way to easily connect your samples and store them, either in alcohol or in the fre ezer, until you can do the work. This means you will need some kind of collector and lots of containers to keep to bees or mites in.
Many researchers use a modified hand-held vacum cleaner to collect bees into a vial, then mark each vial carefully as to colony number and date collected. Don’t put the treatment on there as that may bias you later when counting mites. Store your samples in the freezer, or in a safe place if using alcohol or sticky boards, until later, when you can count mites and record the numbers on the data sheet. One last thought. Make sure you can dedicate these colonies to the experiment for the entire duration of the experiment. Pulling several away in the middle renders the research useless. However, a colony that dies count s.
Now that everything is in place, you’re made all of your decisions and you’re ready to go, recheck all of your notes, plans, maps, diagrams, lists and ideas to make absolutely sure you haven’t forgotten to record something. Too many notes is far bett er than too few. Once samples are counted, add up the numbers and do some simple statistics to see if there is significant difference between treated and untreated colonies. This is important. You must be able to show that your results are not due to c hance alone. We’ll explore the wonderful world of statistics next month.