Navigate / search

How to keep bees like a scientist

As scientific thinkers, we need to re-examine our beliefs constantly and question what we “know” to be true. As we learn, our cumulative knowledge increases, often giving us a better perspective. For example, I recently had a wake-up call that forced me to re-evaluate the idea of the windshield effect.

This article first appeared in American Bee Journal, Volume 160 No 9, September 2020, pp. 993-996.

The windshield effect is a phenomenon that people have noticed the world over. In decades past, car travel from town to town often entailed a good bit of windshield cleaning as the glass became laden with squashed bugs that blocked the driver’s view. In those days, gas station attendants often did this for us as they filled the tank, and many travelers kept a supply of cleaners, rags, and scrapers in the car.

When I was young I couldn’t reach the windshield, so I was given headlight duty. The lenses would get so burdened with dead insects that light barely passed through, and the slop needed to be removed frequently in order to see the road. I can still smell those ammonia-soaked rags and the wormy odor of bug goop that permeated my clothes.

For the most part, we no longer clean car parts between stops, which most of us see as a good thing. But environmentalists around the world are horrified by clean windshields and blame pesticides, urbanization, climate change, and habitat loss for the missing bugs. So last year, when an entomologist told me the windshield effect was nonsense, I was taken aback. Then I decided he was simply too young to remember.

A different view

But several months later, when I was interviewing Dr. John Ascher of the National University of Singapore, I asked him about the windshield effect, and his answer stopped me short. He explained that the relative number of bugs on windshields then and now doesn’t mean much because we have no idea what was in the mix. Did it contain a great diversity of critically important insects or did it comprise several species in huge out-of-control numbers like a plague of locusts? In short, are we mistaking insect “blooms” for healthy and diverse ecosystems? Are we confusing biodiversity with sheer volume?

The truth is we will never know. No one back then thought the bugs were of any long-term consequence, so no one saved the windshield scrapings for posterity. We’ll never know if those legions of road bugs were a good thing or a bad thing because we don’t know what species they were or their relative abundance in those ecosystems.

And other variables can enter the picture, too. In the past, busy roads often traversed agricultural land, but now much of our cropland is set back from heavy traffic. Also, cities and towns frequently spray herbicides along roadways, which suppresses insect populations along long stretches of pavement.

In truth, we probably do have fewer insects than we used to and less biodiversity, too. But the windshield effect doesn’t prove that theory because it’s merely anecdotal evidence.

What is anecdotal evidence?

An anecdote is a story, so anecdotal evidence is story evidence. Right now, I’ll tell you a story about swarms. This spring I had three swarms in the trees in my backyard, and all three swarms selected my one and only top-bar hive as their preferred home and nixed my baited Langstroths. True tale. From this observation, I could easily conclude that honey bees are attracted to top-bar hives, or perhaps they are repelled by Langstroth hives. The evidence for such a conclusion is story evidence and it is useless. Countless factors could have swayed the bees’ decision, so the story is merely a red herring.

You scoff and say, “No one would come to that conclusion from just three swarms.” Maybe not, so let’s take another, one that regularly arises on my website. Someone writes, “I had a colony that was strong and doing great when I checked on it four months ago. But last week when I opened the hive to harvest the honey, there was nothing left but wax moths. Wax moths totally destroyed my colony.”

My turn to scoff. You had a strong colony, didn’t check on it for months, and now the hive is full of moths? Sounds normal. Moths love a weak or dying colony and somehow you provided one. But you can’t conclude moths killed the colony because you have nothing but anecdotal evidence — a story about what you observed. Oftentimes, anecdotal evidence illustrates a truth, but a story by itself cannot prove it.

Thiking like a scientist avoids mistakes. Photo shows a frame of capped brood and attendant worker bees.
If you’re not thinking like a scientist, it’s easy to come to the wrong conclusions.

What about that red herring?

A red herring is a distraction that throws someone off course. The phrase comes from a fictional story, written in the 1800s by William Cobbett, about a boy who saves a hare from the hounds by diverting the dogs with a dead fish, which just happened to be a red herring. Red herrings are common literary devices, often used in mystery novels, to divert the readers’ attention from the true villain, but the term is also commonly used to describe logical fallacies.

In the wax moth example, the presence of so many moths and larvae fouling the interior of the hive diverts you from the true cause of the colony’s demise. We follow the deceptive trail which leads us away from the answer.

Various variables

Both of these stories ignore the many possible variables that could have produced the outcome you saw. A variable is an element, feature, or condition that can easily change. Going back to my swarms, perhaps it wasn’t the architecture of the top-bar hive that attracted the swarms, but maybe it was the amount of sunlight. Maybe it was the odor of the comb or the recently deceased colony that used to live there. Maybe it was the size of the entrance hole, the height off the ground, or the internal volume of the hive that attracted the bees

Sometimes conditions exist that we don’t account for, simply because we had no idea they were important or because we forget about them. In science, variables that are outside the scope of the experiment, but still affect the outcome, are called extraneous variables or confounding variables.

I knew an entomologist who did extensive controlled testing to learn how much of a certain pesticide caused harm to bumble bees. All the colonies received carefully measured doses, and she had an ample number of control colonies. However, in the course of the experimentation, she forgot to account for ambient levels of pesticide that drifted in from nearby fields, a situation that botched the results.

Tapping the scientific literature

Most of us backyard beekeepers are not in a positon to perform detailed scientific experiments on our bees. We usually don’t have enough colonies for statistical rigor, or we don’t have the manpower to make timely changes, take quality notes, and keep detailed records. Furthermore, do we even want to? Science done right is complex and time-consuming and often doesn’t work as a hobby.

Still, it helps to think like a scientist when you are observing your bees and trying to discover what is right or wrong with them. One important skill is the ability to read peer-reviewed scientific papers.

Shredding papers

When I was working on a master’s degree in environmental studies, I was surprised at the amount of time that was accorded to reading and evaluating scientific literature. We spent months and months reading papers that were deemed poor, okay, and excellent by the professors. We would pick them apart, line by line, searching for errors in experimental design or implementation, and shortcomings in logical reasoning and statistical analysis. At first, I saw little wrong with these papers and, because they were peer-reviewed, I didn’t expect to see egregious errors.

But as time passed and we learned what to look for and how to find it, the errors became glaring. We began to wonder how any of it got published. Years later, when I took a master beekeeping course, we were given similar assignments: read and evaluate. Both of these institutions understood that scientific thinking begins with reading and understanding the literature. It requires a thorough appreciation of the scientific method and knowledge of its shortfalls. If a beekeeper cannot evaluate the stuff that’s published, he will be pulled asunder as each succeeding paper claims a different outcome.

Reading science

You can find papers to support any thought. Neonics are killing bees or they’re not. Corn syrup is bad for bees or it isn’t. Varroa must be micromanaged, maybe.

To understand these papers, the beekeeper needs to evaluate the science for himself and decide if the conclusions are valid based on the experiment. Letting someone else do the thinking is a crapshoot. Oftentimes, two people can evaluate a paper and think it means categorically different things. Members of the press and Facebook can be dangerous, too, often determining a paper’s validity based on the title alone or maybe the abstract.

I’m not saying the scientists are at fault. Far from it. Science done well is extremely difficult. Every single step, from stating the hypothesis, to designing the experimental protocol, to implementing the plan, gathering data, and ultimately analyzing the statistics, is fraught with potholes. Scientists must be diligent, but because research is so exacting, those reading science must be just as scrupulous.

Two types of reasoning

Discussions of scientific thinking often center on the differences between the two major types of reasoning. The first type, called inductive or bottom-up reasoning, reaches generalized conclusions from specific instances. The opposite type, called deductive or top-down reasoning, reaches specific conclusions based on proven facts. If that pan-fries your brain, you’re not alone.

Inductive reasoning is often considered unscientific because it’s not based on facts. Instead, it starts with an observation followed by a search for the explanation. Day-to-day human thinking is mostly of the inductive type. In fact, we’re more-or-less hard-wired to think inductively. While that’s not always a bad thing, if we don’t realize its limitations, it can easily lead us astray.

On the other hand, deductive reasoning is associated with the scientific method, and it is used extensively in scientific research. It often begins with a pair of proven facts, and draws a conclusion based on those facts.

When done properly, deductive reasoning will lead to sound conclusions. The trouble is the “when done properly” part. Unknowns lurk everywhere, and we often don’t see them or even know they exist, which means the statements of fact may be inaccurate. If the facts are not accurate, the conclusion won’t be accurate either.

Beekeeper reasoning

Let’s look at a typical example of beekeeper inductive reasoning. Let’s say you kept bees successfully for five years until someone erected a cell tower a mile from your apiary. During the following winter, all your bees died. Since you didn’t change your management protocol or anything else, you conclude that cell towers kill bees.

In this case, you came to a generalized conclusion (cell towers kill bees) based on a specific observation (a cell tower was built near your apiary and your bees died). This type of reasoning is conclusory and fails to take any other circumstances into account. Isn’t it possible your bees died of pesticide poisoning, disease, or parasites? Or perhaps cold weather, starvation, or something you haven’t thought of? Come now! You simply took a wild guess.

Inductive reasoning is faulty when it takes a problem and assigns a cause without any experimentation. Worse, we often blame the first thing that comes to mind, or sometimes the thing that bugs us the most, like that ugly tower.

Deductive reasoning is more nuanced, and starts with facts. For example, if you say honey bees belong to the order Hymenoptera (true) and all Hymenopterans go through complete metamorphosis (true), you can conclude that honey bees go through complete metamorphosis (true). It’s mathematical. If A equals B, and B equals C, then A equals C.

However, if you say John is a beekeeper (true) and all beekeepers get stung (most likely, but we don’t know for sure), then you cannot conclude that John will be stung. Even if we see John acting like an idiot around his bees, we can’t deduce he will get stung because we don’t know if the second statement is always true.

Logical fallacies

I’m not saying backyard beekeepers need to sit down and plot their logical thinking strategies. That’s nearly inhuman. But when we are trying to figure out what went wrong or understand how we can improve our beekeeping skills, it doesn’t hurt to be aware of some of the more common logical fallacies.

You can find lists of logical fallacies online, sometimes 15 or 20 per article or, in one case, 229. Logical fallacies are most often variations on inductive reasoning, traps that are easy to fall into. And fall we do. All of us.

I’ve picked out a few common logical fallacies, just to give you an idea of how irrelevant thoughts can interfere with logic. Be aware that many of these fallacies overlap in such a way that one circumstance can fall under several categories at once.

Flawed thinking

Confusion between correlation and causation is all around us. Remember those wax moths in the dying colony? Wax moths are opportunists, adept at finding and exploiting weak colonies. But just because wax moths and failing colonies often occur simultaneously doesn’t mean one caused the other. Something caused the colony to weaken, and the moths took full advantage of the situation. Did the moths cause a strong colony to weaken? Not likely.

A hasty generalization is easy to make, especially when it provides an easy answer and eliminates future work. The new beekeeper who says, “Based on my very first mite count, I can see my bees are immune,” is bound for trouble.

An ad hominem fallacy finds fault with the person rather than his ideas. “Joe is an electrician, so what can he possibly know about beekeeping?” If you don’t like what Joe says, it’s easy to roll your eyes and say, “Consider the source.”

A beekeeper makes an appeal to authority when he says, “A master beekeeper taught me, so it must be right.” This type of fallacy invokes a person’s title, education, position, or fame as proof that he is an authority and, therefore, correct.

We often like to set up a false dilemma by arguing there are only two positions and nothing in between. A statement like “You’re either a natural beekeeper or you’re not” does not allow for any flexibility or creativity in thinking.

Beekeepers aren’t the only ones to use the alphabet soup fallacy, but they are particularly skilled at it. You simply convince people you are right by using so many abbreviations they can’t possibly follow your (lack of) logic. “Last year my TBH had DWV, BQCV, and PMS all at once. So I decided to use an SBB, a QX, and a UE. Then I treated with OA and FGMO mixed with HFCS. That worked until they picked up AFB, or maybe EFB, not sure, while foraging in OSR.” Sure, whatever you say.

The slippery slope assumes one wrong choice will lead to a cascade of undesirable consequences. “If we keep treating for mites, pretty soon no genetic resistance will remain, and before you know it, we’ll be treating monthly, weekly, and daily until no bees are left, no plants, and we all starve to death.”

Another time-honored favorite among beekeepers is the appeal to tradition. “My grandfather kept bees in skeps for fifty years. If it was good enough for him, it’s good enough for me.” This thinking disregards changes that occurred during those 50 years, including the addition of varroa mites, tracheal mites, small hive beetles, deformed wing virus, Nosema ceranae, neonics, and climate change.

In a non sequitur fallacy, the conclusion does not have any logical connection to the argument. “My bees died because my boss made me work every second Saturday in February.”

Although it’s fun to use anthropomorphism to illustrate a point, it should not be taken too seriously. Anthropomorphism assigns human traits to non-humans, as in “My bees were angry at me for mowing the lawn.” It sounds logical but anger is a human construct, not an insect one. More likely, the noise was alarming to them.

Scientific thinking is better for bees

The list of potential fallacies is endless and often entertaining. Logical fallacies are a good reminder that our thinking can quickly get mired in information, ideas, biases, and emotions that have nothing to do with rational thinking.

When you feel yourself getting pulled into a less-than-logical argument, step back, and see if you can find the false reasoning. And remember, your bees will be better off and your skills as a beekeeper will improve when you think like a scientist.

Rusty
Honey Bee Suite

Thinking Like a Scientist Makes Beekeeping Easier

Comments

Jeff Richardson
Reply

When I was young – very long ago – I remember the ill effects of the windscreen on our firefly (lightning bug in the SE US) population. Fireflies squashed across the windshield would glow for quite a while which saddened me as a child.

These days we have fewer and fewer lightning bugs during our springs and summers. Again I am saddened that I may not be able to share the joy of catching them by the mason jar fulls with my grandchildren – which is a fond memory of my rural childhood.

Rusty
Reply

Jeff,

I think of lightning bugs all the time because they were a significant part of my childhood in the northeast. I’ve heard they are in serious decline and it makes me so, so sad. I can’t imagine a childhood without them. They were the first bugs I ever let crawl on my bare arms!

Ron German
Reply

Rusty

WHEW! What do you think about wearing masks to avoid COVID?😂

I have a three-year-old queen in the better of two hives. Should I requeen this spring?

Ron
Roca, NE

Rusty
Reply

Ron,

Ha! Regarding masks, I think we should separate science from politics. There. I said it.

Regarding three-year-old queens, that’s getting up there. If she were mine, I’d probably replace her.

Kevin
Reply

Slight difference of opinion here. I think politics should take their cues from science rather than separating.

Granny Roberta in CT
Reply

WHEN I had a boss, I almost always had to work EVERY Saturday in EVERY month, and I never had a single bee die on me back then, but that was my favorite example from your list of fallacies. There IS a reasonably logical conclusion from my opening remark—which is that I didn’t have any bees until I retired. 🙂

Also, when we blame the wax worms or deny the varroa problem, we’re probably just trying to make it “Not my fault”. That way we wouldn’t have to do anything about it.

Also, also, if you run an experiment and find that some beekeepers don’t get stung, that just means you didn’t run it long enough. Send those poor beekeepers back into the fray!

Also, also, also, and as usual, I enjoyed this article. Be well.

Rusty
Reply

Roberta,

Great examples of flawed thinking! I should know, most of mine is flawed, I’m sure.

Granny Roberta
Reply

Nonsense. All of your thinking is flawless. And so is mine.

Just for the record, I did NOT put an emoticon (whatever that is) up there. I used a perfectly dignified and respectable colon and closing parenthesis.

Rusty
Reply

😊 Just this minute I learned how to do that in WordPress. My first, and now second: 😜

Granny Roberta
Reply

: )
(and that better not go all new-fangled on me)

Jason
Reply

“I think we should separate science from politics. There. I said it.”
Thank you for saying it!

ET Ash
Reply

So political science as taught in most universities is not really a science?

From my contact here with the economics department it seems a lot of professors and many of the students (reference appeal to authority above) could not seem to discriminate between economics and political science.

Gene.. hanging out on the left coast…

Cal
Reply

Here’s a fun little book: “How We Know What Isn’t So” by Thomas Gilovich.

Rusty
Reply

Cal,

Ooooh. That sounds fun!

Alun Thomas
Reply

Great article, all true; but surely you know that bees can’t hear you mowing the lawn? So they MUST have been angry because you mowed it!

Rusty
Reply

Haha!

Herb Betz
Reply

One of my favorite “correlation does not mean causation” examples.

Folks at a resort community noticed that drowning deaths at the beach go up in direct proportion to the amount of ice cream cones sold. So the solution – ban all ice cream sales to stop drowning deaths.

Drowning deaths in the summer go up because people come to the beach and go swimming. Some unfortunately drown. Ice cream cone sales go up because people come to the beach for vacation and buy them for their kids. Nothing proves that eating ice cream in any way causes you to have a higher propensity to drown. But they are correlated.

And that would be a fine example of mixing science with politics.

In reality, I don’t think you can separate the two because many many many public policies set by political bodies rely on science. Let’s just hope it is not junk science and faulty logic.

Rusty
Reply

Herb,

I’d forgotten that ice cream example, but you reminded me of it. I think I first heard it in a statistics class. It’s a good one.

ET Ash
Reply

In my own experience most folks who think in a scientific manner ALWAYS tend to question what they know or what they think they know. I would guess the common logically fallacy that goes along with this is ‘confirmation bias’.

I will also add that imho doing science and applying science often require folks who think in quite a different fashion. However, some large degree of persistence seems to be required for both styles of thinking.

Gene hanging out on the left coast…

gap
Reply

Re: bugs. When I was very young, my father kept an old squeeze bottle of water in the car. If the bug load got too heavy between gas stops, he would hold it out the window and squirt the windshield while running the wipers. I’m sure this was not a unique solution. During my adult life, however, cars have universally been equipped with more convenient window washers. So we absent-mindedly clean the window constantly while driving, rather than waiting for the bug batter to build up. Thus my counter theory is that we don’t have fewer bugs, we just don’t have to notice them anymore. How’s that for anecdotal/observational logic? Oh, and on some regular road trips during the last 40 years, our windshield was a mass collection screen, on very specific parts of the route, but only during certain times of the year. We always figured they were normal, seasonal, insect blooms.

Re: “Shredding papers.” It’s too bad this type of attention isn’t taught in journalism schools anymore either. 25 years ago, a journalism summer student shared that they were taught in journalism school that getting the “scoop” was more important than accuracy, because “you can always fix the details later if you have to.” In contrast, her mentor at our company had insisted they get the story details right the first time, as a practical application of the old adage “you never get a 2nd chance to make a first impression.” It was an eye-opener to her. If it actually stuck with her, I imagine she’s had trouble finding or keeping a job in that profession 🙁

Ken Rhodes
Reply

I remember two short years ago while traveling up to Missoula for a Honey Bee event featuring Randy Oliver and the guys from the University of Montana. (Yes, I too took the online classes where we had to read peer-reviewed literature). Along the Blackfoot River on I-15 the bugs were so thick during the evening drive, I had to stop a few times to wash them off the headlights and windshield. Your article reminded me of this😅

Your article is so true, especially with regards the social media (Facebook groups for beekeeping). So many will take what is said or printed as the gospel truth when In fact it may have worked for one beekeeper in one particular location. This does not mean it will always work or even that it will work for you, let alone work again for the person who posted it. Many conclusions can be reached but without accurate records, experiments, documentation, etc… how will we know this will happen again?

Often my advice to people is to listen to others but test what they have to say and use this to come to your own conclusions; then expect a different outcome and you will never be disappointed.

I used to give out lots of advice and opinions as a new beekeeper but soon found it best to give out a few suggested scenarios and leave the final say up to the person asking. Frustrating for some wanting a direct answer but safer for me as I really don’t know as much as they think I do (which really is not much).

As always, I love your posts. Thanks

Rusty
Reply

Ken,

This reminds me of me. When I go to beekeeper meetings/events, I usually don’t say a single word. I learn a lot that way.

JWPalmer
Reply

Rusty, I remember years ago the love bugs being so numerous in East Texas that you literally had to clean the windshield every 10 to 15 miles or so. The dead bugs would lie so thick on the highway that passing vehicles would cause them to stir up like a dust cloud.

Thank you for writing this article. Good science is good science. Bad science is just bad.

Jeremy Barnes
Reply

I have long wondered about the windshield affect, in particular, are modern vehicles more aerodynamic and thus insects are swept over them rather than caught up in them. So I went to an antiques motor show and asked some of the owners if they still got insects caught in the windshield and grills, and the answer was a unanimous no. So again, no proof, but it removes one of the variables.

As a child in the late 40’s/early 50’s, my first butterfly collection came from the grille of my father’s car.

Rusty
Reply

Jeremy,

We have a Model T Ford, and it doesn’t get bugs either, although we don’t drive it very far.

Andrew Hogg
Reply

Great article Rusty. And so applicable to much in today’s life. It reminds me of a great meme that shows Santa saying, “I’ve been reading your social media posts and a lot of you are getting science and history books this Christmas!”

Stay well, and thanks for all your great work.

Rusty
Reply

Andrew,

I like that!

Kevin
Reply

This is a great article. I rank it among the best I’ve read in a long time. It does take a long time to learn and develop scientific thinking, but more importantly, it takes curiosity to facilitate that learning.

One additional thought on the windshield effect. Comparisons between the “now” and “then” are made more difficult due to the more aerodynamic design of modern cars where air flows more around cars rather than into cars.

Victor Hall
Reply

Thanks for sharing this, Rusty. Truer words have never been spoken. I taught chemistry for 40 years, and I can say from my experience that students are inductive thinkers.

This should be required reading for all beeks and soon to become beeks. I think that it could ameliorate a lot of comments on social media.

I wanted to add to Herb’s and Kevin’s thoughts. In my opinion, political bodies ofttimes try to USE science to establish a TRUTH and set public policies that conform to an IDEOLOGY rather than science.

There, I said my piece. I do hope that I did not offend anyone as that was NOT my intention.

aDub
Reply

I thought that I’d point out that wax moths do not attack and kill colonies like some aggressive foe. Nor are they the result of “bad beekeeping” (unless you are weakening your colony). Wax moths are a natural part of the hive ecosystem, just like AFB & varroa (now). A strong colony will keep the wax moth population under control. Using chemicals to alter the balance, is IMHO stupid. Change the genetics or environment instead. In the wild, the wax moth has the important job of “cleaning up” after a colony has died, removing all remnants of the old comb and making space for the next one. I would assume that in doing this they also manage infectious diseases that may affect a swarm taking up residence in a space previously occupied by a dead colony since the bees will build completely new comb.

Interestingly, the only person that I’ve ever heard mention the reduced numbers of insects on windscreens and number plates, until now, is me! (maybe it has come back to haunt me like my (fake) “bad day virus”?) The data although, as you say, is empirical, does not detract from the trend, there is no doubt from anyone that I have spoken to that the numbers on cars have diminished. However, I would suggest that a contributor to lower numbers of dead insects, including bugs, may be the improved aerodynamics of modern cars. We should take some cars from the 70’s & 80’s for a spin 🙂 I would bet money that there would still be a massive reduction.

There is without doubt fewer insects, just about fewer everything, with humans a glaring exception! Fobbing off with “insect blooms” is, to my mind illegitimate, unless of course, that bloom has lasted 10s of millennia. To blur the debate, however naïve one considers the contributors, is dangerous and thoughtless – Let no one doubt that we are on a knife’s edge.

Alistair
Reply

Interesting comments about windshield effect. I always wondered whether part of the issue is that cars are far more aerodynamically designed than in the 1960s (when I remember the bugs!), so more of the insects should be pushed around the headlights and the windscreen.

Jeff
Reply

Three points regarding the science discussion above:

1. The change in automotive aerodynamics over the last 50 years,
2. Perceptual bias (find what we expect), and
3. Probiotics solve pesticide poisoning.

Regarding bullet point #3, a recent promotion of a product, based on a peer review of an exceptionally limited quantity of bees/hives, is getting some traction. Before jumping on this bandwagon, let’s see some real science.

Happy holidays all.

Adam Zwickle
Reply

What a great post! I teach a course for new PhD students on scientific methodology and research design and we cover these topics. In fact, it is my love for science that keeps me keeping bees. I often think about my activities from a scientific lens, but you have said this all better than I ever have. Perhaps I should have my students read this?

Rusty
Reply

Yes! Have them read it. Maybe there’s a future beekeeper in there.

Bryn
Reply

I would briefly add that the lack of bugs on windshields, might be due to better aerodynamics of the vehicles. I have noticed how many bugs are splattered on my front license plate, but very few in comparison to my windshield, also how my back windshield is dry, even in a rain storm when driving at speed.

All the best.

Leave a comment

name*

email* (not published)

website

This site uses Akismet to reduce spam. Learn how your comment data is processed.