Monday, April 30, 2012

Splash-Dunk (x2) + Spin-Dry Video

Here's one of my favorite splash-dunk/spin-dry videos.  It was taken at Cranberry Lake in Anacortes, WA.

The dragonfly in this video, most probably a male Paddle-tailed Darner, makes two nice splash-dunks.  It then takes a long, smooth glide toward the water as if it might dunk again.  At the last moment it changes its mind (apparently) and decides to head upward for a spin-dry.  The resulting spin sheds a large number of water droplets.

Check out the post "How Slow Is Slow" to get a feeling for how slowed down this video really is.

Also, recall the following overview that gives a good idea of what's going on in splash-dunk/spin-dry behavior:

Sunday, April 29, 2012

Getting Into A Spot Of Trouble

Yesterday, Betsy and I explored the Hassayampa River Preserve for the first time.  It's a wonderful place, with a great variety of habitats, ranging from a lake, to a stream, to cottonwood groves, and even open desert.

One of the treats along the stream was a thriving population of damselflies, including American Rubyspots.  The photo below shows a rubyspot (right) fending off an aggressive California Dancer.

Notice how the rubyspot flashes the red patches on his wings, which are usually hidden as it rests with its wings closed.  At the same time, it arches its abdomen and opens its appendages, which look like the pincers on an earwig.  The display generally has the desired effect of warding off the dancer; even so, the dancer usually lands just a short distance away on the same twig to await another chance to get the rubyspot out of his territory.

Here's a closer look at the rubyspot's threat display.  Can he actually do harm with those menacing appendages?  I wonder.

The Blue-ringed Dancer

Blue-ringed Dancer (male) at the Gilbert Water Ranch

One of the most photogenic damselflies is the Blue-ringed Dancer.  This male is perched on a rock in a swiftly flowing stream, the Blue-ringed preferred habitat.  I love the intense colors, especially in the head and thorax.

Notice how the wings are held completely above the abdomen; this is characteristic of dancers as a group.  In contrast, bluets tend to hold their wings folded along their abdomen.

Saturday, April 14, 2012

Visualizing Splash-Dunk/Spin-Dry Behavior

Here's a schematic representation of the splash-dunk/spin-dry behavior of dragonflies:

Thursday, April 12, 2012

Extreme Maneuver

Dragonflies are well know for their acrobatic abilities in flight.  Here's an extreme example:

This dragonfly, as you might have guessed, is a male Happy-face (Paddle-tailed Darner).  It was patrolling back and forth along its territory, a walking path in front of some bushes near a water-filled ditch.  Most of my shots show him in normal flight, but in this shot he is apparently making an extreme turn to the right.  I think fighter jets would have a hard time keeping up with this guy.

Here's what he looks like most of the time.  Notice the heavy eyebrows of the "happy face," and the fact that the front two legs are tucked up behind his head.

Tuesday, April 10, 2012

Splash-Dunking Gone Bad: The Sticking Frequency

(This post is a version of an article I published in Argia 24(1): 19-22, the official journal of the Dragonfly Society of the Americas.)

There are many aspects of the recently described splash-dunk/spin-dry behavior in dragonflies (Walker, 2011) that are of interest. In this paper I concentrate on what happens when splash-dunking goes awry and a dragonfly gets stuck in the water. To put this phenomenon in context, I start with a brief overview of some of the general features of splash-dunking and spin-drying.

The Frequency of Splash-Dunking

Splash-dunking is a fairly common event at Cranberry Lake in Anacortes, WA, where my wife Betsy and I do most of our observing. Though the rate of splash-dunking varies from day to day, as one might expect, on a typical day a splash-dunk event is observed every 5 to 10 minutes.

Figure 1 shows data recorded at Cranberry Lake during the 2011 dragonfly season. The upper set of data points shows the clear decrease in temperature during the season. The lower set of data points show the splash-dunk rate in dunks per hour. The average dunk rate is 6 dunks per hour, and the maximum rate is 12 dunks per hour. Though the temperature drops about 20 ˚F during the observation period, the average dunk rate is essentially unchanged.

Figure 1  Temperature (upper data) and dunk rate (lower data) versus date of observation. The straight lines show the trends in the data; namely, a clear decrease for the temperature and no significant change for the dunk rate.

The Dragon Splash

When people see one of my slow-motion videos of darners slamming into the water during a splash-dunk (Walker, YouTube), they invariably remark on the intensity of the splash, and wonder how the wings survive such an impact. The fact that the wings hit the water with some force is illustrated by the shape and size of the splash that is produced.

Figure 2 shows a head-on view of a splash produced by a splash-dunking darner. The darners flying when this picture was taken were primarily Paddle-tailed Darners (Aeshna palmata), though a few Shadow Darners (A. umbrosa) were seen as well. When viewed from this angle the shape and symmetry of the splash becomes apparent. The image shown here is a frame capture from a slow-motion video, and hence of low resolution. Still, it shows the key features of what I like to call the “dragon splash.” Notice the tri-lobed structure of the dragon splash, with a central component produced by the impact of the body, and symmetric side splashes from the wings impacting the water.

Figure 2  The tri-lobed “dragon splash” produced by a darner impacting the water.  Dragonflies typically splash-dunk 1 to 6 times in succession, each time producing an impressive splash.

Dunk Time

When darners perform a splash-dunk, they don’t dillydally in the water. They generally pop right back out in less than half a second. The number of dunks observed for a variety of time intervals is shown in Figure 3. The bar labeled “0.325” corresponds to times between 0.325 s and 0.349 s, the bar labeled “0.350” corresponds to times between 0.350 s and 0.374 s, and so on for the other bars. The average time it takes for a dragonfly to emerge from the water after a splash-dunk is 0.383 s.

Figure 3  Number of dunks versus time spent in the water.  The first bar is for times from 0.325 s to 0.349 s, the second bar for times between 0.350 s and 0.374 s, and so on.

Spin-Dry Parameters

After doing 1 to 6 splash-dunks, a dragonfly rises well above the water and does a spin-dry, which usually consists of 5 rotations and lasts about 0.44 s. Rotation rates have been observed as low as 760 rpm and as high as 1,600 rpm. The average rotation rate for our observations is 1,014 rpm.

There’s a good reason extended spins with many more than 5 rotations are not observed. A complex object in three dimensions – like the body of a dragonfly – has three independent axes of rotation, each of which has its own moment of inertia. Rotation about the axes with the maximum and minimum moments of inertia is stable, but rotation about the axis with the intermediate moment of inertia is not stable. In the case of a dragonfly, the axis of rotation through the wings – which is the axis of the spin-dry motion – is the one with the intermediate moment of inertia. As a result, the spin-dry motion is inherently unstable. In fact, dragonflies pulling out of their spin-dry are often observed to be “wobbling” as they complete their last spin, a sign that the instability is affecting their rotation.

When Splash-Dunking Goes Wrong

Life doesn’t always work out as planned. For splash-dunking dragonflies, this means that sometimes they don’t make it back out of the water. If they can’t become airborne again in half a second or less, they just aren’t going to make it at all. The result is generally death by drowning, though predation may play a small role as well.

Figure 4 shows a male Paddle-tailed Darner (A. palmata) that hit the water about 50 yards from shore and promptly became stuck. We refer to this as a “sticking event.” All sticking events we observed occurred on the first splash-dunk, all happened with air temperatures below 65 ˚F, and all were irreversible.

Figure 4  A male Paddle-tailed Darner (A. palmata) struggling to escape the water after a splash-dunk that didn’t go well.  At this point the darner is close to shore, after struggling for several minutes, and its wing beats are weak.  Just after getting stuck its struggles were much more vigorous, several times getting the dragonfly to the verge of escape.

It’s difficult to watch these gutsy animals struggling to free themselves from the water after getting stuck. They try so hard, come so close to escaping, and continue to struggle for such a long time. The individual shown in Figure 4 struggled for several minutes until – surprisingly – it “paddled” its way to shore right in front of me. I took the opportunity to rescue it and place it on a bush in the sun. After several minutes of drying out it took wing, apparently no worse for wear. I couldn’t help wondering if it would splash-dunk again.

We observed the first sticking event on September 19, 2011, after having observed 90 successful splash-dunks starting back on the 4th of July. As the season progressed, and the temperature dropped, the sticking frequency increased to higher and higher levels. At the end of the season, when the temperature had dropped into the upper 40s, the sticking frequency was a full 25% – one in four splash-dunks resulted in death. The close inverse correlation between temperature and sticking frequency is shown in Figure 5.

Figure 5  Temperature (upper data) and sticking frequency (lower data) as a function of the date of observation.  The inverse correlation between temperature and sticking frequency is evident.

The same kind of behavior was seen in the fall of 2010, before we started collecting data. I remember going to Cranberry Lake one day in late October 2010 when the temperature was below 50 ˚F. I would say as many as 10 darners were stuck in the water and trying to escape at any one time. It was depressing to see them struggling, knowing their efforts were futile.

In Figure 6 we plot sticking frequency as a function of temperature. Notice the nice fit to an exponential decay with increasing temperature. Another way to state this is that as temperature is decreased, the rate of increase in the sticking frequency is roughly proportional to the value of the sticking frequency. In this sense, the sharp rise in sticking frequency seen in Figure 5 is an indication of the dragonflies “hitting the wall” when it comes to their low-temperature flight capabilities.

Figure 6  Sticking frequency as a function of temperature.  The drop-off with increasing temperature is roughly exponential.

Minimum Flight Temperature

Our observations at Cranberry Lake show that dragonflies like the Paddle-tailed Darner (A. palmata) can fly at ambient temperatures as low as 44 ˚F. This is in sharp contrast to a minimum flight temperature of 57.2 ˚F reported for aeshnids (including A. palmata) in Alaska (Sformo and Doak, 2006). In any case, it’s clear that flight at such low temperatures is pushing the envelope when it comes to a dragonfly maintaining the elevated thoracic temperature necessary for the flight muscles to operate efficiently. Sformo and Doak report thoracic operating temperatures in A. palmata of about 97 ˚F.

It’s difficult enough for a dragonfly to maintain the necessary high thoracic temperature when the surrounding air temperature is below 50 ˚F, but the situation becomes much worse when the dragonfly splash-dunks into water. Even though the water temperature was the same as the air temperature at Cranberry Lake (within ±1 ˚F), water drains thermal energy away from a dragonfly at a much higher rate than air. Specifically, Newton’s law of cooling states that the rate of transfer of thermal energy is proportional to both the temperature difference and the thermal conductivity (Walker, 2010). Noting that the thermal conductivity of water is about 23 times greater than that of air, it’s clear that a dragonfly will lose thermal energy rapidly when it is in contact with cool water. Though a number of studies have addressed thermoregulation in dragonflies at high-temperature extremes (May, 1976; May, 1995), less attention has been paid to low-temperature performance, and so far none seem to consider cooling due to contact with water.

Finally, one might wonder whether the increase in sticking frequency with decreasing temperature could be caused by an increase in the surface tension of water, making it harder for the dragonfly to escape (Kuntz, 2012). While this may be a contributing factor, the surface tension increases uniformly by only about 2% over the same temperature range where the sticking frequency increases sharply by over 20%. It seems the most important factor determining the sticking frequency is maintaining the thorax at operating temperature.


The sticking frequency of splash-dunking dragonflies shows a strong inverse correlation to ambient temperature. In fact, dragonflies engaging in the splash-dunk/spin-dry behavior when temperatures are less than 50 ˚F are at significant risk of becoming stuck in the water – which is lethal.

Literature Cited

Sformo, T. and Doak, P. 2006, Thermal ecology of Interior Alaska dragonflies (Odonata: Anisoptera). Functional Ecology 20: 114-123.

Kuntz, R. 2012. Private communication.

May, M. L. 1976. Thermoregulation and adaptation in dragonflies (Odonata: Anisoptera). Ecological Monographs, 46(1): 1–32.

May, M. L. 1995. Dependence of flight behavior and heat production on air temperature in the green darner dragonfly Anax junis (Odonata: Aeshnidae). Experimental Biology 198: 2385-2392.

Walker, J. S. 2010. Physics. Addison-Wesley, 4th edition, pg 555-559.

Walker, J. S. 2011. Spin-Dry Dragonflies. Argia 23(3): 29-31.

Walker, J. S. 2011. Splash-Dunk Analysis, 2011. Argia 23(4): 29-30.

Walker, J. S. Slow-motion videos of splash-dunking and spin-drying can be found on the YouTube channel "thedragonflyguy".