Friday, January 31, 2014

The Whisperer Speaks: Intriguing Aspects of Dragonfly Behavior

On Saturday, I had the pleasure of giving two dragonfly talks for the Sound Waters Conference.  Thanks so much for attending, and making it an enjoyable experience for all.

Here's a brief description of the talk I gave on dragonfly behavior:

My background, which is mainly in theoretical physics, is described below.  I've been studying dragonflies with increasing interest for the last 6 years or so.

Here's the handout for the dragonfly behavior talk.

After the talk, I was asked about the camera that was used to take the slow-motion videos.  They were taken with a simple point-and-shoot camera called the Casio EX-FH25, shown below.  I shoot the videos at 240 frames per second, which gives good time resolution for observing dragonfly behavior.

I was also asked about the Dragonfly Society of the Americas, and their scientific publication, Argia.  You can find more information about the DSA at the website for Odonata Central, as shown below.

The link to this page is

Here's an example of Argia, showing the first paper I published describing the spin-dry behavior of dragonflies:

The Whisperer Speaks: Identifying Dragonflies in Puget Sound

On Saturday, February 1, I gave a pair of dragonfly talks for the Sound Waters environmental conference in Oak Harbor.  These talks were a little more detailed and technical than the ones I've given in the past.  The first talk was on identifying common Puget Sound dragonflies, and the second one dealt with interesting aspects of dragonfly behavior.

Here's a description of the ID talk:

Here's page 1 of the handout I gave out for the identification talk.  It deals with identifying common non-darner dragonflies by color:

Here's page 2 of the handout, which deals with identification of common Puget Sound darners:

Thanks for attending, and participating with many interesting and insightful questions.  You helped make it an enjoyable experience for all of us.

Happy dragonflying!

Wednesday, January 8, 2014

A Hovering Darner, Part III

The previous two installments on this topic focussed on the vertical motion of the darner as it hovers.  We've seen that it bobs up and down with a frequency of about 4.5 bobs per second, with roughly 4 strong wingbeats alternating with 4 weak wingbeats for each cycle of the bobbing motion.  These installments can be found at the following links:

We now include the horizontal motion in our analysis.  I didn't know what to expect when I started, but it turns out that the darner is making a series of "orbits" as it hovers.  Here's an example of an orbital cycle:

The orbits are traversed in the counterclockwise direction, but this is not significant because the orbits would be clockwise if viewed from the other side of the darner.  What is significant is that the darner moves forward as it drops in altitude, and moves backward as it rises.  This is the same regardless of the side from which it is viewed.

Here's another example of an orbital cycle.

The series of bobbing cycles shown below is interesting in terms of the corresponding orbits.  You can see that cycle 2 rises higher, and takes more time to complete, than the other cycles.  Is this difference reflected in the orbits?  Let's see.

First, for comparison, let's look at the orbits for cycles 1, 3, and 4.  As we can see below, these orbits are quite typical in shape.

When we look at the orbit for cycle 2, however, we see something quite different.  As shown below, the cycle 2 "orbit" consists of two loops – a sort of loop-the-loop orbit – with both loops orbited in the counterclockwise direction.  Thus, cycle 2 does have a different type of orbit, after which normal bobbing cycles continue as before.

Finally, I've collected below a number of orbits in a 3x3 grid.  The center square shows the takeoff of the darner, and the surrounding squares show a variety of different orbits, including the loop-the-loop in the lower right square.

What's particularly interesting to me about this figure is that it was drawn by the dragonfly itself.  Each orbit was traced out by the darner during one of its up-and-down bobs.  Nice work, Mr. Happy-face!

Tuesday, January 7, 2014

A Hovering Darner, Part II

The analysis of a hovering darner continues in this post.  Part I can be found at the following link:

Last time, we saw that a hovering darner was actually "bobbing" up and down several times per second as it hovered, a frequency much slower than the rate of its wingbeats.  We also produced a plot showing all twenty up-and-down "bobs" that the darner made in a five-second video clip.

To see the vertical motion in greater detail, we expand the plot and show the flight level of the darner for a period of 1.6 s.  Notice that the bobbing motion is remarkably regular in both frequency and amplitude.  I really hadn't expected to find such a nicely sinusoidal behavior.

The sinusoidal motion of the darner is well represented by the mathematical relation y(t) = A sin(wt), where y is the flight level relative to the average height, w is the angular frequency, t is the time in seconds, and A is the amplitude of motion.  From the above figure, we can see that the amplitude is approximately A = 0.40 cm, and the period of motion is about T = 0.22 s.  From the period of motion, we can obtain the angular frequency w as follows: w = 2π/T = 29 rad/s.

The result for the angular frequency becomes important when we calculate the vertical acceleration of the darner.  To begin, we note that the acceleration a(t) is obtained by taking two time derivates of the position; that is, a(t) = d^2y/dt^2 = –Aw^2sin(wt).  It follows that the maximum upward or downward acceleration of the darner is Aw^2 = 3.4 m/s^2.  This is roughly 1/3 the acceleration due to gravity.

It follows, then, that the darner flaps weakly for about 4 flaps, resulting in a downward acceleration, followed by 4 strong flaps, resulting in an upward acceleration.  This is illustrated in the next figure.

One explanation for the bobbing motion during hovering is that the alternating spurts of strong and weak wingbeats gives the darner a chance to rest on the wing – a little like wind sprints in humans.  Another possibility is that the bobbing motion is a display mechanism – bobbing up and down may increase a darner's visibility to his rivals, letting them know where his territory is, and that he is on guard and ready to attack any intruders.

The last installment in the hover analysis will be posted tomorrow.

Monday, January 6, 2014

A Hovering Darner, Part I

Last summer, Betsy and I spent a delightful afternoon at Beaver Pond, near Winthrop, WA.  It's a medium-sized pond with a tall snag topped by an Osprey nest, and surrounded by a wide flat trail lined with towering Ponderosa Pines and ghostly Quaking Aspen.  The day was sunny and warm, and buzzing with activity of all types.

At one point we sat on a bench to relax, and a friendly male Happy-face Darner (shown below) began to hover at eye level, just inches away from us.  After a few seconds of hovering, he would dart off after a rival, and then return to continue guarding his territory.  He did this time after time, repeatedly giving us a great view.

A male Happy-face Darner (aka, the Paddle-tailed Darner) hovers in front of us, keeping us company.

At first, I wasn't going to take a video of his hovering – after all, I have plenty of good hovering videos from previous years.  But then I noticed something that I had noticed before, but hadn't given much thought.  As the darner hovered, it was "bobbing" up and down.  This doesn't seem too surprising, given that the hovering is produced by wings that are flapping up and down.  What was surprising, though, was that the bobbing was at a much lower frequency than the wingbeats.  Why such a difference, and just what is the bobbing frequency?

With these questions in mind, I decided to go ahead and take a video of the hovering.  The total 5 second video can be found at the following link:

The video was shot at 30 frames per second, and so I made screen captures of each of the 150 frames in the total clip.  I digitized each frame by measuring the distance from a prominent point in the background (a crossing of two branches) to a fixed point on the body of the darner.  This removed effects of the camera moving slightly during the clip.  I measured to the head of the darner, to his thorax, and to the tip of his tail – all three cases gave essentially the same results.  The results given here were recorded for distances measured to the darner's head.

The graph below shows a plot of all 150 data points for the 5 seconds of hovering.  The "cluster" of points is the hover itself, and the points going off to the right show the darner's takeoff to challenge a rival.  Notice that this is a plot of vertical position versus horizontal position.

To see what's happening during the hover, we expand the cluster of points, and plot them as vertical position versus time.  The result is shown below:

The first thing we notice in this plot is that the darner is indeed bobbing up and down as it hovers, and in a fairly regular way.  In fact, it's clear there are about 4 - 5 "bobs" per second, as compared with the wingbeats, which are at the rate of 35 - 40 per second.  The second thing we notice is that the average height of the hover is decreasing with time at a steady rate of about 0.2 cm per second.  This is something I hadn't noticed in real time as I observed the hover.

The analysis of the hover will continue in a subsequent post.