Pollinator Observation Experiment: Who Picks What?

Here is a prediction you can test this afternoon. Honeybees and bumblebees will largely skip the red flowers in your garden and pile onto the blue, purple, white and yellow ones. The reason is not preference in any fuzzy sense: bees are trichromatic with receptors tuned to ultraviolet, blue and green, and they see long-wavelength red poorly, so a red flower is a dark smudge to them. Butterflies and hummingbirds, which do see red, take those flowers instead. A pollinator observation experiment with a timer and a notebook turns that textbook claim into your own data in about thirty minutes, and the counting method matters more than the watching.

Key Takeaways

• You will run timed five-minute counts at three or four flower patches, record visits by pollinator group, and divide by the number of open flowers to get a fair comparison.
• The concept is pollinator ecology: flower colour, shape, scent and nectar position sort visitors into predictable groups, a pattern biologists call pollination syndromes.
• Best for ages 6 to 13. Younger kids tally; older ones handle the per-flower normalisation and the order-rotation that controls for time of day.
• The usual failure is comparing raw counts. A patch with forty open flowers will get more visits than one with six, and that tells you nothing about preference.
• You observe. You do not catch, net, jar or touch anything. That is a rule, not a suggestion.

What You’ll Need

Item	How much	Where to get it	If you don’t have it
Flower patches in bloom	3–4, clearly different in colour and shape	Garden, balcony pots, park, verge	A florist bunch in jars outside works, with flowers of known types
String or four sticks	Enough for a 50 × 50 cm square per patch	Shed	Pace it out and mark corners with pebbles
Timer	1	Phone	Kitchen timer
Tally sheet	1 printed grid per patch	Paper and ruler	Notebook with columns
Pencils in two colours	2	Pencil case	One pencil, different symbols
Phone camera	1	Pocket	Field notes and a guide book
Folding stool or cushion	1	House	Sit on the ground
Sun hat, water, insect repellent	As needed	—	—

Time: prep 15 min, counting 30–40 min, analysis 15 min Cost: free Mess level: none

Do the Experiment

Step 1 — Choose patches that differ in one obvious way

Pick three or four patches whose flowers differ in colour and shape: for example a lavender or catmint patch (small, blue-purple, tubular), a daisy-type flower (flat, white or yellow, open centre), a red or orange tubular flower, and one scented pale flower. The USDA Forest Service describes these groupings as pollination syndromes, with bee flowers tending to be bright white, yellow, blue or ultraviolet with nectar guides and a landing platform, and butterfly flowers bright including red and purple with a narrow tube and a wide landing pad.

Step 2 — Mark a square and count the open flowers

Lay out a 50 × 50 cm square over each patch and count exactly how many open flowers are inside it. Write that number down before you start watching. This is the step almost every version of this activity misses, and it is the one that makes the result mean something, because visits scale with how much food is on offer.

Step 3 — Watch one square for exactly five minutes

Sit at least half a metre back, stay still, and tally every visit where an insect lands on a flower. Sort visits into groups rather than species: honeybee, bumblebee, other bee, hoverfly, butterfly or moth, beetle, wasp, other. If one bee visits six flowers in your square, that is six visits. If it leaves and comes back, that is a new visit.

Step 4 — Rotate the order and repeat

Do each square once, then do them all again in reverse order, then if you can a third round in a shuffled order. Pollinator activity changes through the day with temperature and light, so without rotation you cannot tell a flower effect from a time-of-day effect. Note the time, temperature and whether it is sunny for every round.

Step 5 — Work out visits per flower and compare

For each patch, add up visits across all rounds, divide by the number of open flowers, then divide by the number of five-minute blocks. You now have visits per flower per five minutes, which is a rate you can compare across patches. Build a small table with one row per patch and one column per pollinator group.

What you should see: On a warm, calm, sunny day between about 10 a.m. and 4 p.m., a healthy lavender or catmint square will draw bees steadily, often one visit every few seconds. Daisy-type flowers pull in hoverflies and small solitary bees. The red tubular flowers will usually get very few bee visits and may sit empty for a full five minutes, while butterflies work them. Expect enormous variability: one cloud, one gust of wind, and the numbers collapse. The honest outcome of a single afternoon is a pattern, not a proof, and three separate days turn it into something you can defend.

What’s Actually Going On

Flowers advertise and pollinators shop. The advertising is colour, shape, scent and the position of the reward, and because different animals perceive and handle those differently, plants have evolved packages aimed at particular groups of visitors. Briscoe and Chittka’s review of insect colour vision lays out the sensory basis: most bees are trichromatic with photoreceptors peaking in ultraviolet, blue and green, with nothing tuned to long-wavelength red. So a pure red petal reflects almost nothing a bee’s receptors respond to. It is not invisible, but it is low-contrast and uninformative, and bees do not search it out. Birds do see red well, and hummingbird-pollinated flowers are typically scarlet, orange or red, odourless and funnel-shaped with deep, ample nectar, which is the USDA Forest Service’s description of the bird syndrome.

The nectar guide is the part most people have never noticed. Many flowers carry patterns that converge on the centre, and some of those patterns are strongly ultraviolet and therefore invisible to us but conspicuous to a bee. The Forest Service lists nectar guides as present in both bee and butterfly syndromes and absent in bird and beetle ones. A child who learns that a plain yellow flower may be covered in markings they physically cannot see has learned something more useful about perception than any amount of “animals see differently” ever conveys.

Two honest caveats. First, counting visits is not the same as measuring pollination. An insect that lands, drinks and leaves without contacting the anthers has done the plant no good at all, and some visitors are outright nectar thieves. Visitation is a proxy, and a good one, but it is a proxy. Second, garden plants bred for show are often poor food. Doubled ornamental varieties, the ones with ruffled extra layers of petals, frequently have their nectaries and anthers replaced by petal tissue, so a spectacular flower can be a desert. If one of your patches gets near-zero visits from everything, check whether it is a double cultivar before concluding anything about colour. The Forest Service’s own pollinator gardening advice is to include plants native to your region, plant them in clumps rather than singly, keep a succession of blooms from early spring into late autumn, and eliminate pesticides where possible.

How to Explain It to Your Kid

Ages 5–8: Flowers are shops and bees are shoppers

Tell them each flower is a little shop selling sugar, and the colours and shapes are the signs outside. Some signs are easy for bees to read and some are not. Their job is to be the person who counts customers: one tally mark every time a bug lands. Then compare the patches and ask which shop was busiest. The idea that different customers go to different shops arrives on its own.

Ages 9–12: Per flower, not per patch

Introduce pollinator and the problem of a fair comparison. If one patch has forty flowers and another has six, the big patch wins automatically, so you divide. Have them do that division themselves and watch the ranking sometimes flip. Then introduce the red result and the reason behind it: bees see ultraviolet, blue and green, and red is nearly off their scale.

Ages 13+: Syndromes, proxies and controls

Give them the framework. Pollination syndromes are a hypothesis about convergent evolution, not a law, and plenty of flowers are visited by groups they “should not” attract. Have them design the controls themselves: rotation order for time of day, equal observation length, per-flower normalisation, and multiple days for weather. Then push on the weak point: visitation is a proxy for pollination. Ask how they would measure actual pollen transfer, and let them discover that it is hard.

The question to ask: “Our red flowers got almost no bee visits. If we painted a blue flower red with food dye, would the bees stop coming to it, and what would that tell us?”

Why Teaching This Early Pays Off

The concept-specific payoff is a correct model of animal perception. Children are taught that bees and butterflies like flowers, which is true and useless. The far more interesting fact is that a bee lives in a different visual world, one with ultraviolet in it and almost no red, and that flowers are shaped by that. Once a child has that, a lot of biology becomes tractable: why some flowers open at night, why moth flowers are pale and strongly scented, why bird flowers have no smell at all.

The methodological payoff is the denominator again, and the rotation. Dividing visits by the number of open flowers is a small act with a big idea inside it, and the step of changing the order in which you observe is a child’s first controlled experiment in the real sense. Alfieri and colleagues found that guided discovery helped learners while unguided exploration did not, and this is a clear case: a parent pointing out “wait, that patch has more flowers” converts an afternoon of watching into an afternoon of reasoning.

Keep the claim honest. Thirty minutes of counting produces noisy data and a plausible pattern, and the most valuable thing a child can say at the end is “we would need more days to be sure”. That sentence is worth more than the result.

A Pollinator Observation Experiment and What Parents Should Know

Best age	6 to 13. Sitting still for five minutes is the real age limit, and some six-year-olds manage it better than some ten-year-olds
Supervision	Adult present, mainly to enforce the no-catching rule and to manage distance from the flowers
Real hazards	Bee and wasp stings. Stay at least half a metre back, move slowly, no swatting, and no bare feet on clover. If your child has a known sting allergy, carry the prescribed adrenaline auto-injector and do this from a window or at a distance instead. Also sun, ticks in long grass, and wasps around food and drink
Don’t	Do not catch, net or jar any insect, and do not handle bees even gently. Do not use flowers that have been sprayed with insecticide, and do not spray anything to “see what happens”. Do not pick the flowers you are counting mid-experiment

When It Doesn’t Work

• Almost no visits anywhere. Check conditions first. Below about 13 °C, in wind above a light breeze, in rain or under heavy cloud, pollinator activity drops sharply. Early morning and late evening are also quiet. Try 11 a.m. to 3 p.m. on a warm still day.
• One patch gets nothing at all. Before blaming colour, check whether it is a doubled ornamental variety with no accessible nectar, whether it has been sprayed, or whether it has already finished flowering and the “open” flowers are spent.
• The counts are wildly different between rounds. Normal. Insect activity is patchy in time. More rounds and more days is the only fix, and reporting a range rather than a single number is the honest presentation.
• Your child cannot tell a hoverfly from a bee. Nor can most adults, and it does not matter much. Use broad groups, note “unsure” as its own column, and photograph anything interesting for later identification.
• A bee visits every flower in the square in one go. That is six visits, not one visitor, and it is a real behaviour called flower constancy. Record visits, and note separately that it was a single individual. Both numbers are interesting.

Take It Further

• Colour as the only variable. Set out identical jars of sugar water on cards of six different colours, including red and ultraviolet-reflecting white, and count landings. Keeping shape and scent constant isolates colour in a way flowers never can. Place them away from doors and seating.
• Morning against evening, same patches. Count at 9 a.m., 1 p.m. and 7 p.m. on the same day. Bee visits usually peak midday; pale scented flowers may do nothing until dusk, when moths arrive. That is the night-flower syndrome showing up in your own numbers.
• Submit what you photograph. Upload clear photos to iNaturalist with location and date on, and track which observations reach Research Grade, which requires an accurate date, coordinates, a photo or sound, a wild organism, and more than two thirds of identifiers agreeing at species level. Pair it with running a real bird count and submitting the data, which uses the same discipline about effort and honesty.

What to Watch For Over the Next 3 Months

• Week 4: They notice a flower colour and predict its visitor before anything lands. That prediction, right or wrong, is the model working.
• Month 2: They ask about something they cannot see — ultraviolet patterns, or how a bee finds a flower on a grey day. Both are real research areas and good places to let curiosity run.
• Month 3: The follow-up worth encouraging is planting. If they want to put in flowers that actually feed pollinators, use natives for your region, plant in clumps and aim for continuous bloom, and start with seed bombs and which natives actually sprout. To see how a flower moves water and nutrients internally, dyeing a white flower and tracing the plumbing is the companion build.

Frequently Asked Questions

Do bees really not see red?

They see it poorly. Bee colour vision is trichromatic with sensitivity peaking in the ultraviolet, blue and green, and nothing tuned to long-wavelength red, so a pure red petal gives them little signal. They can sometimes find red flowers by shape, scent or ultraviolet reflectance, which is why “never” would be too strong.

Is it safe to sit near bees with a young child?

Generally yes. Foraging bees on flowers are not defensive and sting mainly when trapped or squashed. Keep half a metre back, move slowly, wear shoes and avoid sugary drinks nearby, which attract wasps. If your child has a diagnosed sting allergy, observe from a distance and carry the prescribed auto-injector.

How many five-minute counts do we actually need?

For a classroom-quality result, three rounds per patch on each of three different days, all in similar weather. One afternoon gives you a pattern worth talking about; three days gives you something you could put on a poster without overclaiming.

Should we count wasps and flies as pollinators?

Yes, in their own columns. Hoverflies are significant pollinators, some wasps carry pollen, and the Forest Service lists separate fly and beetle syndromes with their own flower types. Lumping everything as “bees” is the commonest error in amateur pollinator counts.

---

About the author

Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.

---

Sources

1. USDA Forest Service. “Pollination Syndromes.” Celebrating Wildflowers. https://www.fs.usda.gov/wildflowers/pollinators/What_is_Pollination/syndromes.shtml
2. USDA Forest Service. “Gardening for Pollinators.” Celebrating Wildflowers. https://www.fs.usda.gov/wildflowers/pollinators/gardening.shtml
3. Briscoe, A. D., & Chittka, L. (2001). “The evolution of color vision in insects.” Annual Review of Entomology, 46, 471-510. https://doi.org/10.1146/annurev.ento.46.1.471
4. iNaturalist. “What is the Data Quality Assessment and how do I use it?” iNaturalist Help. https://help.inaturalist.org/en/support/solutions/articles/151000169936-what-is-the-data-quality-assessment-and-how-do-i-use-it-
5. Alfieri, L., Brooks, P. J., Aldrich, N. J., & Tenenbaum, H. R. (2011). “Does discovery-based instruction enhance learning?” Journal of Educational Psychology, 103(1), 1-18. https://doi.org/10.1037/a0021017
6. National Research Council. (2012). A Framework for K-12 Science Education: Practices, Crosscutting Concepts, and Core Ideas. National Academies Press. https://doi.org/10.17226/13165