Saturday, 2 July 2016

What's a bird's favourite colour?


In this experiment, we looked at which coloured feeder birds choose to feed from. Colour clearly matters in the lives of birds. In many species, a mate is chosen because of their plumage, with sexual dimorphism and different colours between the male and female present in a large number of birds.

Colour is also used for camouflage, and although the males nearly always have a different plumage to the female in Anatidae (ducks and waterfowl) this is because the female alone incubates the nest, and their dull plumage is needed for camouflage.

Birds are some of the most fascinating and captivating living things on this planet, and that is because of their fantastic and colourful plumage variations.

In this report, we’ll look at how to carry out this experiment, and ways to cancel out variables, as well as the results and why they might be so, including an in-depth look at the eye of a bird and how that might influence the choice that they make.

Pre-Investigation research

Any experiment requires research before hand, to ensure that what you are doing is possible, and to find out the details. Because it’s summer, we chose to use mealworms as one of the foods, because these are especially popular at this time of year, with adults rearing chicks. This would ensure that we had maximum visits to the feeders, and therefore more reliable data on which to make conclusions.

We also researched the possibility of comparing the results for this to a similar moth experiment. However, this would have been far too complex to carry out, and it wasn’t prime moth-trapping season. Furthermore, financially, it would have been too far fetched, because a moth trap that is safe for the moths costs in excess of £120, and we would need five of those!

Before we began the experiment, we asked ourselves the key questions: Location? How long should each feeder be outside for? How many days should the experiment last before the birds learn that you can get the same from all the feeders? How should we position the feeders?


The main aim of this investigation was to ascertain which is the favourite colour of garden birds, and the reasons as to why this might be. It is a highly practical experiment, and one without a theory. The experiment is also highly complex, because you need to cancel out all the variables. Why do this experiment? First of all, it is easy to do, but proves something that is as yet unknown. Also, it is relatively easy to do, and it is the epitome of “citizen science” due to the fact it can be carried out by you at home; it’s cheap and you don’t need a Hadron Collider! Furthermore, this experiment is “one of a series” that could lead on to other experiments. Once you have the results for birds, why not try moths, dogs, rats, grasshoppers or Axolotls! Otherwise, you could see if the result varies between species: is an Ortolan Bunting’s preference different to a Yellow-billed Sap-Sucker. Also, this experiment could  “help” your day to day life; perhaps yellow bird feeders weren’t so great after all: maybe try green. It could also help the birds. If a certain colour was found to repel or deter birds, then it could somehow be used in crop-growing farms to lower the volume of seed eaten by the birds, and stop the farmers poisoning them, be that intentionally or unintentionally.  

The control variables in this study were:
•    Type of feeders used
•    Space between feeders
•    General location of feeders ie. in the same tree, yard or feeding station
•    Scale used to weigh the feeders
•    Kind of bird food at the same time
•    Volume of bird food in each feeder
•    Number of feeders used
•    How long the feeders are outside for

The independent variable was the colour of the feeders, and the dependant variable was the number of visits to each feeder.


The method for this experiment is a rather complex one, for it involves cancelling out many variables so that the results are viable and reliable. Also, it may vary depending on the time of year.

1.      Purchase five identical feeders-these do not need to be high quality, just enough to last two weeks.

2.      Paint four of these in different colours: red, yellow, dark blue and green. Avoid painting the feeding ports, because if ingested, the paint could seriously harm the welfare of the birds. Allow the paints to dry. Keep one feeder CLEAR; this is the control.

3.      You will also need a feeding station or an appropriate area in which to place the feeders, such as at similar positions in a tree.

4.      Soak five sets of 26g of dried mealworms (a total of 130g) in 50ml of freshly boiled water per 26g for five minutes. Then drain.

5.       Fill the five feeders with the soaked mealworms so that they weigh 100g, feeder and soaked mealworms included. This is only about a centimetre above the top of the feeding port: don’t be alarmed!

6.      Then, place the feeders outside.

7.      The feeders will need an “installation period” of 24 hours, so that the birds can become comfortable.

8.      The following day, ideally dawn or just after, record the birds for a set amount of time: one hour.

9.      This is the day split into periods: Morning: 6am-11am  Afternoon: 11am-4pm  Evening: 4pm-8pm

10.  Do an “evening watch” also for one hour.

11.  Note the total number of visits to each bird feeder in a tally chart, as well as the species that visit.

12.  After three days of recording, swap the food for a standard seed mix, and fill each feeder so that it weighs 325g, feed and feeder included (that equates to 275g of feed).   Weigh the mass change of the mealworm feeders.

13.  Record the feeders with seed for three days as well, and do as above with the mealworm feeders (ie.  count and weigh…)

14.  If you find that the feeders are going down very quickly, then weigh the mass change at the end of each day, and refill the feeders to the set mass.

15.   Ideally, repeat everything in a second location, or repeat in the same location.

The method is fairly straight forward, but it does require a certain level of commitment. A way to collect even more data (we had 1064 visits just by watching) would be to set up a motion-sensitive trail camera on the feeders, but this does put up the price of the experiment, and seems not needed. Also, you need to ensure that the birds do not become reliant on this technicolour food source, because removing it so suddenly could harm the birds. To stop them relying on the experiment, you need to make sure you only leave the feeders in one place for three days. Alternatively, you could do your bit and start feeding the birds!

Safety Assessment

You need to be careful when working with hot water for soaking the mealworms; use oven gloves if you are soaking the mealworms in a container that conducts heat. Also, be patient with the mallet, and ensure that your other hand is out of the way before you strike the wood atop the bottom section of the feeding station.


By using two different food types, you are cancelling out the variable of the food attracting the birds, rather than the colour,  just so long as you use the same food type at once.  Also, if you record the birds at different times of day, you cancel out the variable of time of day. By recording the birds from more than one location, you can show that the result is the same, and “your” birds were not anomalies, or prove that there are differences in preference between individual groups of birds.


·         5 identical feeders

·         4 cans of spray paint-blue, red, yellow, green

·         Masking Tape (for covering feeding port when painting)

·         1 feeding station with 5 hooks

·         Mallet (for driving station into ground)

·         Protective piece of wood: this goes on top of the screw fitting in the bottom section of the feeding station when you use the mallet to put it into the ground, otherwise you could damage the screw fittings.

·         1 kg bag of standard bird food seed mix

·         3x 100g bag of dried mealworms

·         Set of scales

·         Notebook for records

·         Binoculars (optional)


So what does our results show?

This is the overall table of results, and it shows that green is the most popular overall, followed by clear, blue, yellow and red.

This above graph shows each species’ preferences overall. This is the data collated from all locations on all days. Woodpigeon may be irrelevant: the woodpigeon only briefly gained access to two of the feeders by using a waterbath as a step! We can’t give a conclusive answer for robins, goldfinches and magpies, because there are not enough visits.

We have tested our results by sheer number: 1064 visits. As a result, our results are fairly reliable owing to the fact that we have so many visits to remove anomalies.  However, our results may lose some of their reliablility because some feeders, even on the feeding station, were more exposed than others, but field observations seem to suggest that they didn’t seem to mind which feeders they visited. However, we overcame this issue by moving the feeders one space clockwise at the end of each day. Weather could have affected the results, but between each day, the results were very similar. Also, there was nothing that we could do about this.

House Sparrows (Passer domesticus), quite clearly favour blue, but green and clear are still popular. The section below explains this in more detail, but the blue, green and clear (which has brown food in it) are in the middle of the spectrum, which the birds are seemingly attracted to. However, yellow and red are least popular, because these are nature’s warning colours.

Goldfinches (Carduelis carduelis) favoured red and yellow, the main colours of their plumages.

Starlings (Sturnus vulgaris) were slightly less picky, but blue still won over, probably for the same reasons as the House Sparrow. However, the higher numbers for red and yellow suggests that the Starlings learnt that you could get the same food from all the feeders more quickly than the House Sparrows.

Great Tits and Blue Tits favoured the blue overall, which not only mimics their plumage, but it too fits with the “middle of the spectrum” theory.

The above graph shows that the choice made by the birds does not change at different times of day, and there is a peak in feeding in the morning and evening (which is already known). By recording the birds at different times of day, you cancel out the variable of time of day.


We also weighed the mass change of each feeder after three days, and this graph shows the average % mass change for each feeder after three days. There is one issue with relying on mass changes, and this is because one or two Starling visits remove a far greater weight of food, but the same number of House Sparrow visits (which are smaller birds, and have shorter visits, because they are usually flushed by Starlings) remove much less than the Starlings. However, overall, it does follow a similar trend to the visits. By recording both types of data, we have a greater amount of information, and have shown different ways to see which is a bird’s favourite colour.

Notes from the field

“One male great tit seemed to peer at all of the feeders from above before actively choosing the green one, unlike some of the House Sparrow records which involved individuals crashing into the feeders.” This is where the green top to the clear feeder is an issue.

“When I repositioned the feeders, the House Sparrows still chose blue and green.”

“The reason behind there being so many clear visits with the mealworms may be because a) they look brown, and that is near the middle of the spectrum and natural camouflage b) it resembles their natural food.” Blue and green still had more visits, however.

“Lots of the birds perch atop the feeding station before actively choosing a certain feeder.”

“Next time, we should ring birds to see if there are individual preferences.” This idea is good, but you need a 5 year training programme to get a licence from the BTO!

How does this compare?

This experiment is not a new idea, but few have actually got a conclusive result. The RSPB carried out a summer investigation in Shropshire, in which blue came first, followed by dark blue, dark green, burgundy, silver and red. There were twelve feeders in this experiment, and the RSPB admitted that the result was not that reliable because woodpigeons had gained access to some of the feeders, meaning that food levels went down more quickly in the accessible colours…

Further studies showed that in summer, blue was the most popular, with dark green also high-up, but silver was the most popular all year round. However, in none of these experiments was a “clear” feeder used, which is the control. For summer, though, the results are similar. It would be interesting to carry on this experiment to see if the time of year influenced the results.


Of course, this result is not totally conclusive, but it could be applied to a business with further exploration. For example, a colour that deters birds could be used to stop birds from flying into windows, or could be used on farms in some way to stop birds eating grain or seed. Furthermore, from a business point of view, the production of green bird feeders could be the way forwards; colours which the birds are comfortable with could be used for bird-watching clothing, hides, accessories (such as scopes or binoculars) or general landscaping. Also, green feeders could be manufactured.

Furthermore, wind turbines could be painted red and yellow to stop fatal collisions with birds; equally, you could paint them green/blue because it would stand out more, because it’s in the middle of the spectrum.

Birds' Vision, Science, and Science Summary

Human Vision:

Sight is a highly complex, and it uses the brain from the front of the head,  to the back of the head, where vision is actually processed. To produce what we know as sight, the eyes need to capture information, and ‘send’ this through the optic nerve to be processed by the optic lobe. The optic lobe is one of the four major lobes, of the cerebral cortex, of every mammals’ brain; this is based at the apex of the brain towards the back, which is then surrounded by the parietal lobe, as well as the temporal lobe. This is the ‘visual processing centre’ of all mammalian brains. The brain further incorporates other information, such as that of sensory stimuli, that is used in conjunction with sight . This really is the very basics of visual interpretation, through light information, or that of photons. A misconception is that you see with your eyes. In actual fact, you see through your eyes, and with your brain; eyes are a mere organ of collection. And like with many different sensory organs, the brain is the organ that interprets this information: the ears, eyes, mouth and nerves collect it. So we have looked at the basics of mammalia visual perception,  but how does this relate to our feathered, dinosaur-derived friends?

Bird’s Vision:

The bird’s eye most closely relates to and functions like that of a reptilian eye. This is unlike that of a mammal’s eye;  it isn’t spherical, and the flatter shape of the bird’s eye allows for a wider field to be in focus. Around a bird’s eye, we find a circle of bony plates. This is named the sclerotic ring surrounding the eye and holding it rigid. However,  there is an improvement over the reptilian eye, which is also found in mammals,  that the actual lens of the eye is pushed forward, into the front of the eye, which in turn increases the size of the image on the retina. Due to the surrounding bones in the eye, the bird cannot move the ball by itself. However,  there are anomalies. The placement of the eyes varies between bird species, and this decides what type of vision  the bird will perceive. For example, birds that have their eyes placed to the sides of their heads have a greater field of view, and this is usually found in birds that are predated.  Meanwhile, those with eyes on the front of their heads- much like that of mammals- have that of binocular vision and can estimate distances with precision.

Figure 1

 Figure one shows the field of view of an owl, a predator, to a pigeon, often lower down in the food chain, maybe a secondary predator. As we can see, the pigeon does have a perception of distance, but only minor, and this is where the two eyes overlap each other, in input. However, it has a near 360 view.

Mammals have eyelids, to maintain the health of the eyes and maybe to rid the pupil of debris. However, birds  and reptiles, have a “nictitating membrane.” This lubricates the eye, and then a third concealed eyelid (that sweeps horizontally across the birds eye) makes a third membrane in the eye of a bird. The nictitating membrane also covers the eye, acting as a contact lens to the bird. It’s often used by many aquatic birds, when swimming, to protects the oculus from any damage. The eye is also cleaned by tear secretions from what is known as the lachrymal gland and is then protected by other oily substances that are created and dispersed by the Hadrian glands.

Small birds, such the Eurasian robin, are effectively forced to be diurnal because their eyes are simply not large enough to give the bird adequate night time vision. However, despite this fact, many birds migrate at dusk and at night (such as the Redwing) but this does result in collisions with protruding attributes to the scenery, such as windmills; this can be fatal. Birds of prey on the other hand are diurnal, even though their eye mass to body mass isn’t particularly low. Their ocular vision is optimised to give the very best of spatial resolution rather than light gathering, so they also don’t function well in lower light levels. Also, owls that aren’t diurnal fill the night-time hunting niche. Common redstarts and the European robin sing much earlier in the day than birds of the same size. This may be down to the fact that the smaller bird, such as our robin, has to start the day earlier due to weight loss overnight. Over a typical night time period, 5-10% of a bird’s mass may be lost and sometimes peaking at over 15% in harsh winter nights. To regain this mass, they have to take advantage of the light, and so feed at earlier times of the day, and further exploit the fact that not as many predators will be out at this time.

But what birds can make sense of night time vision? Well, nocturnal birds, these have eyes optimised for high visual sensitivity, with vast corneas relative to eye length, were as that of diurnal birds have longer eyes relative to that of the corneal diameter, or pupil. The highly dilated pupil allows for more light to enter in, and is therefore far more sensitive. For example, you can see the corneal diameter increase and decrease in humans, relative to what sort of light is emitted into their eye.

In any eyes, there are two types of photoreceptors: rods and cones. Rods contain the visual pigment rhodopsin and are better for night time vision, as they are sensitive to small quantities of light. Cones, on the other hand, pick up certain wavelengths of light, and therefore specific colours, so they are more important to colour orientated animals, such as that of birds. Most birds are tetra chromatic, processing four different types of cone cells. In some birds, the maximum absorbsion peak of the cone cell responsible for the shortest wavelength of light, which extends to ultra-violet light. This makes them U-V sensitive.

Birds’ light perception: 

There are four spectrally distinct cone pigments and these are all derived from the protein opsin. Each neuron in the glandular layer processes different colours and therefore different wavelengths. They process this information with a number of photorecepting cells, and may in turn trigger a nervous impulse to relay the information across and along the optic nerve.The more intense the light is, the greater the excitation of the cone, or multiple cones, and the brighter the light is perceived. 

Iodeosin is a pigment that is most abundantly found in birds. This is a cone pigment that absorbs light at near 570 nm. This is around the spectral colour region that is occupied by red and green sensitive pigments, but this is most common in primates.  In aquatic birds however the pigment has seemingly shifted its absorption peak downwards to near 543nm. This is presumably an adaption to a blue aquatic environment.

The information from one cone alone is highly limited; the cell alone can’t tell the brain which very wavelength of light caused the excitation. A visual pigment, such as the ones stated above, may absorb two wavelengths equally, but these may have varying energies.

Why would one bird be attracted to one colour, in this case the colour of the feeder? We have shut down any variables that we think may obscure our results (for example, the orientation of the feeder, and the food within the feeder). We have also had this feeder in three different areas, allowing us to rule out anomalies. For example, a blue tit we found, nesting near one of the positions to the feeder, seemed to always feed from the blue, green and clear stations; never the yellow or red. To rule the fact out that this is personal preference, we observed other blue tits, in other places, and we still found these result. This backed up our idea that this bird wasn’t an anomaly, but that it represented the species as a whole. However, why would this be? There are many trends in our data that suggests that birds go for the colour that looks most like themselves, but this doesn’t explain birds such as the woodpigeon. There are definite links between plumages and colour preferences. For one, we can look at the robin: it visited the red and yellow stations, colours similar to the Eurasian robin’s breast. The blue tit: visiting the blue and green feeders, once again similar to their plumage. Great tit: visited the green and blue stations, with the rare yellow sighting-look at its feathers, blue, green, and yellow! But how does this work for the more dull species of bird. The house sparrow, a very common visitor to our feeder (but once more common) is a bird mostly composed of various shades of brown.

Threats and Toxins

One thing in nature that is quite common, is that birds, and other animals, flee when they see colours of warning. In nature, fire, a danger, is the red/yellow colour. Outside the UK, the same applies to many frogs and snakes, to deter predators. This colour is evolutionarily locked into the birds to stop them from going near such a danger. Some of these colours are in our feeders (red and yellow) and so birds may be deterred by that colour due to the fact that they may think it’s a threat to that bird’s health. This could be one reason for them not going for the brightly coloured feeders. But yellow still had 102 visits, and red still had 84 visits. Why was this? As the birds became more familiar with the feeders, they eventually found out that the red and yellow feeders gave the same goods as the blue, green and clear feeders. Anecdotal evidence also suggests that Starlings learnt this quicker, because after less than an hour, the Starlings were beginning to visit the red and yellow feeders as much as the blue, green and clear. To make the results more reliable, we should have stopped recording sooner, so the birds didn’t learn that the feeders were, essentially, the same.

If we were looking at invertebrates, we could think that flowers, of a bright colour, are more likely to be a better source of nectar. We may see that there is an abundance of bugs and other insects on red and yellow, as that gives a sense of security, and evolutionarily, they will go for these colours.

Natural Camouflage

One of the other main reasons that green is so very popular is due to natural camouflage. These colours include grey, brown and green; these all attract birds that are slightly more nervous. This factor is overriding, due to the fact that these colours are known to the birds, and lead to an increase in bird visitors to the feeders, due to this overall familiarisation. However, one other thing we do know is that humans, and their brain, are attracted to the centre of their visual colour spectrum, and therefore green/ blue. However, the centre of a bird’s colour perception is in fact blue and green. So why were there so many visits to the clear feeder? One of the elements that plays a large part in the clear feeding appliance is that it has a very earthy green top and bottom to it, and this may be one of the deciding points to its popularity. Also, the clear feeder is filled with brown food (seed and mealworms). This support the idea that they are attracted to the centre of their colour spectrum. Looking to the diagram above, we see that the centre of the spectrum for the bird is this rather brown, dirty green, and this is the same colour that was presented on the top of the clear feeder, and inside. This may have been a flaw to investigation, as they are attracted to this colour. I now conclude that the initial hypothesis is right; birds, like humans, are attracted to the centre of their colour spectrum and therefore a blue/green/brown colour.   


In simple, the eye works like this: light passes through the cornea, the clear, dome-shaped surface that covers the front of the eye. The cornea then refracts this light. The amount of light coming into the eye is controlled by the iris, the coloured area; this regulates the size of the pupil, and therefore the amount of light that enters the eye. Behind the pupil is the lens; this is the clear, photosensitive tissue that contains the photoreceptive cells that convert light impulses to electrical impulses. These signals are then processed by the brain, and the optic nerve allows for a passage for the signals to pass through. We see with our brains; our eyes are merely collectors of information.

In any eyes, there are two types of photoreceptors: rods and cones. Rods contain a hugely important pigment named rhodopsin and is far better for night vision, as they are sensitive to small quantities of light. Most birds are tetra chromatic, processing four types of cone cells, with the fourth being ultra-violet.

The possible reasons behind the bird’s choosing one feeder, were quite numerous. Birds go for the feeder that  looks most like themselves, or a mate. The results also showed that bird species which have sexual dimorphism or where the male and female have differing plumages, colour mattered more. House Sparrows (who fall into this category) hugely favoured green, blue and clear, but Starlings (in which both genders are identical) seemed to have a fairly even spread of choice.

Red and yellow were highly unpopular choices. This could have been due to nature’s colour warnings; they are often red and yellow.

Green, and other natural colours were popular, possibly because they are natural camouflage. These colours are abundant in their lives, and so allow for the birds to feel secure around that feeder.

Birds, like humans, are attracted to centre of their colour spectrum: green/blue/”browny”. Blue, clear and green were popular choices in the selection; in fact green turned out to be the most popular colour to birds.

According to our data, birds choose the feeder that is the colour of the middle of their spectrum, and are deterred by other colours, possibly because of them being the colours employed by nature to warm predators.

What would we do differently?

Our experiment wasn’t perfect, and if we were to repeat the experiment, there would be some things that we would change. One of these is the “clear” feeder. This was the control, but it was topped with a murky-green lid, and had a similar coloured base. This may have affected our results, because to any bird flying in from above, the “clear” feeder would, in fact, look green. Also, the clear feeder actually contained brownish food, so from birds from the side, it would look brown. We could improve this by making the clear feeder white.

Also, it would’ve been better to have used scales that were more precise than to the nearest five grammes, because that could have made the difference, as the “average % mass change” graph shows.

Furthermore, we should, ideally, repeat the entire thing completely twice more, and present data as averages. This would make the results far more reliable because you have ensured that the results were not a one off.

To get an overall favourite colour (slightly anthropomorphic) we should repeat the experiment at different times of the year. Also, we would get an answer for more species if we carried out the experiment in winter, where there are more Starlings, woodland breeding birds are present in gardens (such as robins and the tit species) and there are more finches because they migrate (like Siskins) or that they are in woodland in spring and summer for breeding.

One other factor to consider is the weight loss of mealworms due to them drying back out again. To make the mass change graph more reliable, we should have a feeder with soaked mealworms in, then see what  the mass change is without any birds feeding from it. Although the feeders are all the same, and proportionally it would not matter, some may be in more direct sunlight than others, for example, and the water in the mealworms would evaporate more.


Collins Bird Guide (2nd edition)

Chris Packham’s Back Garden Nature Reserve

RSPB Birds: Their Hidden World (Peter Holden)

RSPB Nature Watch (Marianne Taylor)

“Birdwatch” website and magazine

With Thanks to Mr. Ramshaw and Ms. Chadwick for offering us this fabulous opportunity.

Ed Thurlow and I are through to the National Young Scientist of the Year in March, and we won the "Best Communicators" prize as well in the South West!


  1. Well done in Young Scientist of the Year, and the presentation on Points West this evening!

    1. Hello,
      I was amazed that we won too, and the finished project should be up here soon; it's got the overall data set and other conclusions. It was just a shame that Points West didn't show the far more interesting natural history side to the experiment, rather than the planes, the conclusions about which are not certain yet.
      Thank you! Keep reading,