How Does a Roller Roll

I have read many articles over the years about why a Roller rolls.  But, I do not recall ever reading one about the aerodynamics the Roller needs to master in order to roll.  From a practical standpoint it strikes me that the aerodynamics is an important subject.  After all, if we understand the aerodynamics a bit it might just tell us something about how to breed a better Roller.

Today both video and high speed photography are readily available of Rollers performing.  If you know what to look for these tools are a great help in understanding what a Roller does when it rolls.  But, if you do not understand aerodynamics you can watch these tools as long as you wish and you are not at all likely to ever see the important factors.  For instance I have heard that the roll is initiated by the Roller sticking its tail straight up.  Yes, during a roll a Roller does stick its tail straight up.  But that has nothing at all to do with how the bird generates the angular momentum in the roll.

Airplanes were designed after birds in general.  A major key to making an airplane or making a bird fly in a stable manner is to place the center of gravity and center of lift in proper places with respect to each other.  In both birds and planes the major lifting surface is the wing.  A minor lifting surface in both cases is the tail.  Either a airplane or a bird can rotate on three axis.  These three axis are called the pitch, roll and yaw axis.  For purposes of understanding a bird rolling only the pitch axis is important so I will ignore the other two. 

If you draw a line from one wing tip to the other wing tip rotation around this line is a pitch rotation.  This is the rotation a Roller makes when it rolls.  Modest changes in the pitch axis are made in an airplane by adjustment to the angle of the tail control surfaces.  In a bird the whole tail is moved up or down a bit.  In either a plane or a bird simply adjusting this control surface is able to cause the object to fly a loop providing it has enough air speed at the start.  But, it will not cause the rapid spin we see in a Roller.

There are well understood design criteria for airplanes that lead to stable flight.  If you violate these designs you will have a plane that is either very hard or impossible for a human to fly.  These same design criteria apply to birds.   An important design criteria to worry about is where the center of gravity happens to be relative to the center of lift.  The center of gravity of any object is simply that point from which you could hang the object on a string and it would not rotate nose down or nose up.  The center of lift is that point on a flying object where you could treat all lift as simply being a vertical vector.  In order to have an airplane or a bird that fly's in a stable manner, the center of gravity must be slightly forward of the center of lift.  As you move the center of gravity back towards the center of lift, the flying object becomes less and less stable in the pitch axis.  Very small corrections with the tail control surface, leads to a very large pitch axis response.  If you move the center of lift in front of the center of gravity, stable flight is impossible.  The flying object will roll in the pitch axis.

In an airplane, a pilot who wishes to stay alive worries a lot about where his center of gravity is relative to his center of lift.  There are some controls a pilot has, to move the center of gravity forward or backwards.  He can adjust how the plane is loaded and which fuel tanks are used.  He can seat heavy passengers close to the wing.  A bird is stuck and does not have the ability to move weight forwards or backwards.  A crop full of food moves the birds center of gravity forward and the bird can not shift other weight to compensate.  However, a bird can do something an airplane can not do.  A bird can rotate its wings forwards or backwards at the shoulder to adjust the center of lift, relative to the center of gravity.  This ability to rotate at the shoulder is very important for a bird; if it could not do this, it would have a hard time flying.  Birds simply must be able to make such compensations to accommodate full versus empty crops.   A full crop would call for a more forward wing angle than an empty crop.  A Homer burns a significant amount of abdominal fat during a race moving its center of gravity forwards, from where it was at the start of a race.  So during the race, the Homer needs to be able to swing its wings slightly forward to keep the center of gravity and center of lift in the correct relationship.  A hen can be carrying an unlaid egg one hour, then lay the egg the next hour and need to fly again even though the center of gravity has been shifted.  The hen needs to shift her wing angle slightly forward to compensate for laying an egg.  There is nothing in the slightest unnatural about this ability.  All birds have it.

When a Roller is going to do the first flip it may tilt its body very slightly up from level flight and flare its tail.    At that point its tail is straight out behind the bird. Picture 1 of a Roller right at initiation of the roll shows this clearly.

 

 

Picture 1 also shows the wings rotated forwards from normal flying position.  In normal flying position the leading edge of both wings is nearly a straight line rather than a V as shown in this picture.  This picture and the rest of the pictures in this sequence were all of the same bird during 70% of one initial rotation.

Then the Roller starts a wing flap.  During this wing flap it keeps both wings rotated forwards at the shoulder from the normal flight position.  This moves the center of lift forward of the center of gravity.  The net result is the bird is not in stable flight.  Rather, the lift from that single wing flap starts a violent backwards rotation.  Think about hanging a one foot long ruler from a string tied on the six inch position.  The six inch position is the center of gravity.  Now, if you tied a second string on the ruler at the five and one half inch position and jerked up sharply what would happen?  Obviously the ruler would rotate violently about its center of gravity.  That five and a half inch point is the equivalent of the center of lift.  If you watch flying Rollers and know what to look for you can see this forward swing of the wings for that first wing beat at the start of a flip.  If you watch slow motion of a Roller the wings swinging forwards is a lot easier to see.  What you can not see without slow motion is what the tail is doing during establishment of the angular momentum needed for the flip.   Picture 2 clearly shows the tail is still extended straight out behind the bird.  Yet the bird is already starting to be in a head up attitude.

Picture 5 shows the bird just a little farther along in the first flip of the roll.  The wings are still fully extended over the back and the tail is now fully lifted to reduce drag.  At this point all the angular momentum needed to roll is established.  However, drag forces during the roll will wear away this angular momentum.  So the bird will continue to flap its wings about one time per flip as long as it wishes the roll to continue.  It will also keep it tail high as long as it wishes to continue the roll.  To stop the roll the bird simply extends its tail to normal flying position and does one wing flap with the wings swung slightly behind normal flying position.  This will cancel the stored angular momentum of the roll.
 

What do these pictures tell us about what the best Roller should look like?  One obvious thing is the shorter the birds body is the less angular momentum it needs to generate to allow a given rotation speed.  So a short backed bird should be able to roll faster than a long backed bird.  Likewise a smaller overall bird has less mass thus requires less energy to establish the needed angular momentum.   Flying Rollers typically are small pigeons.  I suspect breeders have done a good job of selecting for proper size pigeons simply by selecting for good performance.  All birds have the ability to swing their wings forwards at the shoulder to put them in position to initiate a roll.  But not all birds are willing to roll.  So the breeder is mainly selecting for an attitude on the birds part that rolling is desirable.  All the breeder can select for is parents that have this attitude based on their performance and hope the attitude is passed to offspring.
 

Some may still feel the tail is important in establishing the angular momentum regardless of what photography so clearly shows.  Well, we all know a bird can fly without a tail.  In fact they can fly remarkably well.  So I took a young roller that had been flying a month.  This bird was doing a lot of single flips and occasionally doing short rolls of three or four flips.  On day one I pulled the two outer tail feathers and released it to fly with the kit.  There was zero change in its performance.  It could still flip just as good and just as often with only 10 tail feathers as it could with 12.  Each subsequent day I  pulled the two outer tail feathers.  I saw no change in the speed of this birds flips as I pulled tail feathers.  Nor did it perform less as it lost tail feathers.  On the day I pulled the last two tail feathers the first couple of times it performed it lost a lot of altitude.  Up until that point it would generally gain altitude during a one or two flip performance.  But after it had adapted to zero tail feathers, which only took less than five minutes flying time, its performance was just as good as it had been six days before with a full tail.  It turned over just as fast and it performed just as often and it no longer lost altitude.
 

On one occasion I have seen one of my birds do a fair flip without even doing a wing flap.  The birds had been flying for some time and were getting tired.  They were doing a lot of gliding.  One of them, during a glide, simply rotated its wings well forward from normal glide position.  The result was a single flip.  It was not a very fast flip.  But it was a well controlled flip.
 

Genetics

 
It seems clear from the photos that any pigeon that can fly is capable of rolling.  All they need to do is rotate the wing forward a bit at the shoulder and the roll is automatic.  Thus the genetics we need to put in the bird are not structural changes.  They are aimed at making the bird enjoy rolling.  The bird needs to want to roll or it is not going to roll.
 

Entrikin’s 1972 Journal of Heredity paper indicates rolling is a multi gene trait.  My observations support this idea.  I started with Joe Quinn stock.  These birds did not perform at all unless they were flown every single day for months.  All of these birds would do a single flip occasionally if flown long enough.  But even two year old birds did not really roll.  The very best of them might turn over three or four times rarely and most only single flipped.  Also the frequency of performance was very low.  Quite often I would watch a kit of 20 birds and see no more than one bird flip in ten minutes.
 

This year I got some stock from a strain that has been bred strictly for performance.  These birds will start to do flips within a month of flying daily.  Some are doing flips while still squeakers.  They progress much more rapidly than the Quinn birds.  From the first tail rides to the first flip is generally only two or three days.  They do good tight flips and quite often by two months of flying are doing triple and quadruple flips.  They also perform at a higher air speed than the Quinn birds.  The Quinn birds need to really slow air speed down to perform.  The high performance birds will perform at normal air speed and routinely exit a single or double flip at higher altitude than they had when they entered the sequence.  Frequency of performance is orders of magnitude higher than Quinn birds.
 

To date crosses between the Quinn birds and high performance birds have given performance no better than pure Quinn birds.  They have now been flying for two months and are not even doing tail rides yet.
 

Quinn states in his CD on Rollers that it took as much as six generations of back crosses to pure Rollers to put the roll back in his birds after an outcross to a nonperforming breed.  Lee Faecking has told me this fits with his experience also.  Rolling is complex.  A lot depends on how often the birds are flown as well as how long they have flown.  Birds flown daily perform more than birds flown only two or three times a week.  The same bird will perform more as a two year old than as a one year old.  It has often been noted that two siblings can differ greatly in performance, even in inbred strains.  It is also a lot harder to judge rolling performance than it is to look at a crest or some other show trait and judge the quality of the crest or the other show trait.  I suspect that some of the many back crosses needed to recover the roll is simply that roll is hard to judge and also due to performance variation in individual birds that are quite alike genetically.  Six generations of back crosses would indicate roll is quite a lot of genes.  Probably over half a dozen at least.  On the other hand when you think about the difficulty of judging rolling and the individual variation, my guess is a half dozen genes is more like the maximum number.  It seems clear that rolling is a multi gene complex just like most physical traits in pigeons.  The data also indicates that some of these genes are codominants.
 

My observation has been that Rollers are quite calm birds relative to many breeds.  I have to wonder if calmness is not an important part of good rolling performance?  A bird that felt stress during its first attempts to roll, as a flighty bird might, would naturally resist performing.  As I feel performance is wholly voluntary and the bird must enjoy it calmness and a low tendency to get excited or stressed may be important.
 

Rollers also seem to have a low degree of homing ability.  Attempts were documented to train a variety of breeds to home in the Hollander years of the Newsletter.  Rollers were the poorest homing breed tested.  Even with considerable training few would come home from 25 miles.  Last year I flew some birds from May 1 until about October 1.  I culled 20 of these birds in early October by cutting their bands off and releasing them 15 miles air distance from my house.  Only one of these birds came home and it took that bird several days to return.  It was a three year old bird.  Homer guys tell me that they would expect 100% return if they did such a test with Homers.  I think Rollers may have been selected for non homing ability.  It is highly desirable for Rollers to fly close enough to the loft so they can be seen most of the time.  I live in a pretty heavily wooded area.  So I can only see my birds fly when they are close to the loft.  If they go more than 1/8 mile horizontal distance from the loft they are going to be behind trees.  I can see them about 75% of the time when they are flying.  By contrast, when I flew Homers they were often out of sight for an hour or more.  Selection for poor homing ability probably induces Rollers to fly much closer to the loft so they do not get lost.

~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.