From the Literature: On Diving Spiders and Physical Gills

I love Twitter!  Since I’ve started using it, I’ve learned more about a wider variety of scientific studies than I ever did before.  It’s a great source of science news!  This story absolutely exploded on Twitter a couple of weeks ago.  I can see why it was so popular though: it’s a super cool story!  Aquatic spiders + special webs for holding air underwater = SCUBA spiders!  This story has it all.  It also happens to be closely aligned to my own area of expertise, so it’s time for another From the Literature!

This adorable little guy is the diving bell spider, Argyroneta aquatica:

diving bell spider

Diving bell spider. Image from /14obspider.html?_r=1&smid=tw-nytimesscience&seid=auto

This spider is one of the only spiders in the world that spends its entire life underwater, from egg to adult.  Like its aquatic insect relatives, the diving bell spider has a number of adaptations that help it survive in water.  Its respiratory adaptations are especially interesting.  See that sheen along the abdomen of the spider in the photo?  That is an air bubble held in place by specialized hairs that trap air against the body when the spider surfaces.  Underwater, the spider can use that air store to breathe.  In addition to carrying the air film, this species does something truly spectacular: they build special silk webs underwater, fill them with air that they carry from the surface, and then use the air stored in the silk balloon (the diving bell) to breathe!  However, neither the air film carried by the spiders on their bodies nor the air contained within the diving bell are enough to completely satisfy the oxygen requirements of the spiders, so they depend on occasional surface trips to survive.

Scientists have known that these spiders use the diving bell as a sort of SCUBA tank for a while, but many questions remained.  No one knew what the oxygen conditions inside the bubble are, how much gas exchange occurs between the air inside the diving bell and the water outside, or how effectively the diving bell acts as a physical gill for the spiders.  (Please see my post Better Breathing Underwater in Aquatic Insects for a complete description of how physical gills work, why oxygen flows from water into an underwater air bubble, and why these air bubbles do not last forever.)  These unanswered questions drove researchers Roger Seymour and Stefan Hetz to look into the oxygen dynamics of the diving bell more thoroughly.  In the process, they discovered some exciting things about both the spiders and the unique air containers they build!

The researchers did several different things for their study.  First, they measured the volume of empty diving bells and used specialized oxygen detecting equipment to measure the oxygen levels both inside the bells and in the water surrounding them.  This information allowed the researchers to use mathematical equations to determine the potential rate of flow of oxygen from the water into the air bubble.  They also measured the oxygen inside the bubble while the spider was inside to calculate the rate of flow of oxygen from the diving bell to the spider, the spider’s oxygen consumption.  Using these two pieces of information, they could then show how effectively the diving bell acted as a physical gill, that is how long a spider could stay submerged when using a diving bell.  The researchers also measured the oxygen levels within the diving bells when spiders voluntarily returned to the surface  to reveal the oxygen level at which the respiratory requirements of the spiders were no longer met by the diving bell and they were compelled to replenish it with fresh air from the surface.  Finally, the pair measured the metabolic rates of the spiders directly using a respirometer, a machine that measures the amount of carbon dioxide released and/or oxygen absorbed by a biological organism.

Based on the result of all these measurements, Seymour and Hetz revealed several interesting things about the spiders and their diving bells.  First, they showed that the larger a diving bell, the more effectively it acted as a physical gill and the longer the spider could remain underwater.  Not only do larger bells contain more oxygen to begin with, but the flow from the water into the bell as the spider consumes oxygen is greater in larger bells than smaller.  Also, the larger spiders, those with greater oxygen requirements as measured with the respirometer, produced larger bells.  The authors further showed that diving bells easily provided the entire oxygen requirement of inactive spiders in warm, stagnant water (i.e., water with low dissolved oxygen) for over a day.  If the spiders moved around, built or cared for cocoons containing eggs  (cocoons are stored inside the diving bells!), ate prey that they captured, etc, then they had to return to the surface more often.  If they remained mostly still, the diving bell more than adequately provided their respiratory oxygen requirements for long periods of time.  Clearly the bells are acting as highly effective physical gills for the spiders!

Seymour and Hetz also demonstrated that the spiders stay within the diving bells until the oxygen drops d to 5-20% of the original level.  At this point, the bell apparently provides insufficient oxygen for the spiders and they return to the surface to collect air to replenish the diving bell.  The spiders also appeared to enlarge their bells if their oxygen demands increased or if the dissolved oxygen levels in the water dropped.

A few important implications are suggested by the results of this study.  The observations and measurements reported in the study were made in rather unfavorable conditions: warm, still water with low dissolved oxygen.  If diving bells are enough to meet the respiratory needs of the spiders for over a day under poor conditions like these, then spiders may be able to stay underwater nearly indefinitely in cooler and/or flowing waters.  This is important for several reasons.  If you happen to be an aquatic spider, you represent a nutritious meal to other aquatic animals such as fish, amphibians, and large insects.  Minimizing your trips to the surface, and thereby minimizing the attention you draw to yourself, is desirable.  It likely requires a lot of energy to make trips to the surface and back as well, so staying underwater as long as possible may help the spiders conserve energy.

diving bell spider in diving bell with cocoon

Diving bell spider in diving bell with cocoon. Image source:

The spiders are also apparently able to respond to their environmental conditions and adjust the properties of their diving bells to match!  Seymour and Hetz observed their spiders enlarging bells under several conditions.  Spiders that captured prey enlarged their bells and added air to them before they started eating.  Spiders with cocoons also enlarged the bells as the broods inside developed, so the parents may be able to compensate for the increasing oxygen demands of their offspring as they develop by modifying the bell.  Apart from demonstrating how effectively the diving bell acts as a physical gill, I think the most exciting result from this study is that it reveals how these spiders intentionally modify their environment in response to their changing needs.

In summary, the diving bell does act as a physical gill for diving bell spiders as scientists have long proposed.  These bells allow the spiders to stay underwater for a very long time, and the spiders can adjust the bells to match their oxygen requirements and the dissolved oxygen levels of the water.  However, even under the most favorable conditions, the air contained within the diving bell will eventually need to be replenished, so the spiders will always depend on air from the surface and must have access to the surface to survive.

Super cool, right?  Aquatic spiders are amazing enough on their own, but spiders that build little air balloons to breathe underwater are infinitely more interesting!  Because I think it’s helpful to see it, I’ll end this post with a YouTube video (not my own) of a diving bell spider building a diving bell.  (I recommend turning the sound off – it’s got obnoxious music).  Enjoy!

Literature Cited:

Seymour RS, & Hetz SK (2011). The diving bell and the spider: the physical gill of Argyroneta aquatica. The Journal of experimental biology, 214 (Pt 13), 2175-81 PMID: 21653811


Unless otherwise stated, all text, images, and video are copyright © 2011

Behavioral Responses of Damselflies to Storms

Fountain Creek Park

The pond where I did my research at Fountain Creek Regional Park, CO.

Last week I talked a bit about how weather affects odonate behavior, my favorite topic in biology. Today I’ll go over the study I did to look at these weather related behaviors more closely. Like the little study I did that focused on damselflies and weather in my first college ecology class, this study was done at the wetlands in Fountain Creek Regional Park outside Colorado Springs, CO.  This research was actually part of my undergrad senior thesis!

pond at Fountain Creek Park

My study site. It extended from the cattails on the left side out to the end of the log in the water and from the dock (not visible) to just beyond the log.

I had to work the summer I collected my data, but I went to the wetland most days after work and recorded observations from 4-5PM.  I plopped down on the dock with all of my weather measuring equipment and watched the damselflies in a 5.5 square meter area along the edge of the pond for an hour.  I divided the hour into 5-minute periods and recorded weather data (wind speed and direction, temperature, light intensity, barometric pressure, relative humidity, and whether it was raining or not) for the first minute of each period.  I then spent the remaining 4 minutes counting the number of damselflies that flew within my study area.  Part of that area was filled with cattails and the rest was over open water as you can see in the photo.

Having spent 14 years of my life in Colorado Springs, I can tell you one thing with certainty: in the summer it rains nearly every day between 4 and 5PM.  This meant that I was out watching damselflies during the exact time the storms were blasting over Pike’s Peak and ripping across the plains.  I would sit there watching these phenomenal storm clouds rolling straight toward me with fantastic speed.  Guess who got rained on A LOT that summer?  Me!  I also got hailed on, was sandblasted in high winds, and was once driven running the half mile back to my car when the lightning got a little too close.  However, the clouds moved so quickly (they have to build up a ton of momentum to make it over Pike’s Peak’s 14,115 feet!), the storms didn’t last long, usually 30 minutes at most.  During that time, the weather would transform from hot, sunny, and still to cold, windy, and rainy in the span of a few minutes.  It would usually rain, sometimes very hard, for 10 minutes or so.  Then the storm would suddenly be over and it would become sunny, warm, and still again.  This whole series of events would take place during my hour at the pond.

Now most sane people go inside during storms.  Rain in Colorado is incredibly cold and the storms can be quite powerful with a lot of lightning.  Call me crazy, but I loved curling my whole body into my enormous rain jacket and getting rained on.  I was rewarded for my insanity too because I got to see some things that very few odonate people get to see.

First, I learned that there was a rather distinct pattern of behaviors that was associated with the weather patterns I observed.  The damselflies were most active in sunny, warm, still conditions, the typical weather central Colorado experiences during the summer.  They flew readily into and out of my study area, hunting, looking for mates, mating, and laying eggs.  As soon as a storm approached, you’d see some pretty interesting things.  As the clouds moved in and it became darker and cooler, the number of flights the damselflies made decreased so that fewer individuals flew during a counting period.  As the wind picked up, the activity decreased even further.  Flight activity ceased altogether if it started to rain.  Of all the many hours I spent at the pond, I saw only a single damselfly flying while it was raining, and it was during a very light rain when the sun was still shining.  Most interestingly to me, the damselflies would start to leave the pond when the weather deteriorated sufficiently.  They were displaying pond abandonment behavior.  However, as soon as the storm was over and the sun came back out, the damselflies would return to the pond and resume their normal activity as if nothing had happened at all.  It was fascinating and I am so happy I got to see this behavior!

pond at Fountain Creek Park

My pond at Fountain Creek Park during a light storm.

The flight activity of the damselflies at Fountain Creek Regional Park was clearly affected by the weather, but I was interested in knowing which of the seven weather parameters I measured were contributing to the flight activity I observed.  I used a statistical procedure (multiple regression for those interested) to determine that light intensity, temperature, wind speed (but not direction), and whether it was raining or not were the weather parameters most closely associated with the flight activity that I recorded.  Of these, light intensity showed the greatest association, followed closely by temperature.  Essentially, the brighter and warmer it was, the more damselfly flights you see.

(Brief aside: Remember how I said last time that I didn’t agree with that Russian scientist who thought that barometric pressure was a major player in shaping odonate behavior?  My results didn’t indicate that barometric pressure had any effect.  This coupled with the fact that the Russian didn’t even measure barometric pressure in his study makes me skeptical of his results.)

So four weather parameters were important.  The statistical test confirmed what I’d observed visually, that damselflies flew more readily in good weather than in poor weather.  “Good” conditions were warm and sunny with little or no wind while “bad” conditions were cold, rainy, windy, and dark.  I definitely observed pond abandonment behavior.

The most important question is this: what does all this mean?  I think my data suggest two things:

pond at Fountain Creek Park

My pond at Fountain Creek Park, right after a storm.

1) The damselflies might be able to pick up on cues in the changing weather that alert them that a storm is approaching.  Think about a damselfly, those big wings on a scrawny little body.  If you’re a damselfly, it could be physically dangerous for you to be out in a storm.  Being blown into the vegetation or the water could be deadly, heavy raindrops could impart a significant blow, and evaporative cooling could cause your body to cool down so fast that you can’t escape if the weather gets worse.  Better to leave the pond before a storm than risk getting caught exposed in one.  I think storms are dangerous to odonates, so the pond abandonment behavior that has been so often reported might be a means of protecting them from harm during bad weather.

2) Pond abandonment behavior might be related to roosting behaviors.  Consider these ideas: Damselflies roost in sheltered areas away from the water at night.  Storms usually result in a drop in the light level and temperature, which are the same things that happen as it gets dark at night.  Damselflies disappear from the water before it starts to rain.  It is therefore quite possible that pond abandonment behavior and simple roosting behaviors might be the same thing: odonates returning to their overnight roosts when it gets dark and cools down.  It is likely also advantageous for damselflies to seek shelter during storms, but this could be a secondary benefit, something they gain by completing a behavior that has nothing to do with protecting them from storms.

Are odonates using weather cues to abandon ponds before storms?  Or are they simply returning to their roosts because it’s getting dark?  Are storms dangerous to odonates?  These are some of the endless new questions I had after I finished this project and would like to answer.  I had intended to study this behavior in more depth in grad school, but then I decided to attend grad school in Arizona.  Colorado’s clockwork storms are perfect for studying these behaviors.  Arizona’s wildly unpredictable storms are not.  So, I changed my focus to the water bugs and have studied them ever since.  I will go back to my beloved odonates someday though!  I also decided a while back that my damselfly study was actually pretty unique and could make a real contribution to the scientific literature on odonates.  Ten years after I wrote my undergraduate senior thesis, my data was published.  If you’d like to read more about my study, look at some pretty graphs and whatnot, the citation is listed below.

I am dealing with some heavy things in my personal life at the moment, so I have no idea what I’ll do for the next few posts.  I’m going to let myself be driven by whims for a week or two.  I hope you’ll all check back to see where my whims take me!

Paper citation:

Goforth, C. L.  2010.  Behavioural responses of Enallagma to changes in weather (Zygoptera: Coenagrionidae).  Odonatologica 39: 225-234.


Unless otherwise stated, all text, images, and video are copyright © 2011

From the Literature: Tracking Dragonfly Migrations

It’s about the time of year for the dragonflies to start moving south!  I’ve already gotten several reports of big migratory swarms headed south from several locations across the eastern and midwestern U.S. and I expect many more – the season has just begun!

A few posts back, I discussed a paper that described mass migratory swarms in dragonflies.  While the authors presented several unanswered questions and got the ball rolling on understanding how and why these swarms form, there has still been surprisingly little done in this field.  As I suggested in my last post, this might have to do with the ephemeral nature of these swarms.  The vast majority of swarm observations are “in the right place at the right time” sorts of observations and it’s extremely difficult to predict exactly when and where a swarm will form and/or travel.  Depending on their location, any given dragonfly researcher might only see one or two mass migratory storms in his or her whole life!  This is clearly a very difficult topic to study, and most accounts of swarms have been buried in the scientific literature.  That means that there is very little information about dragonfly swarms freely available to the public.  I think this is a sad state of affairs, thus today I’m covering another scientific paper on dragonfly migrations.  This one is really fun!

(Okay, okay – I think it’s fun, but I’m also a huge bug geek…  You can form your own opinion!)

dragonfly with transmitter

A darner with its transmitter attached. Photo by Christian Ziegler and taken from sciencenow/2006/05/ 11-02.html?etoc&eaf

In 2006, one group of researchers decided to answer one of the big unknowns: where do these migrating dragonflies go?  The group, headed by Martin Wikelski of the Department of Ecology and Evolutionary Biology at Princeton, had noticed that though many insect species had been documented migrating, the ultimate destination and migration strategies of many of those species remained unknown.  So, they decided to track swarming dragonflies.  How did they do this, one might ask?  With radio transmitters of course!  Check out the photo to the left.  That’s a green darner (Anax junius), the most common swarming dragonfly, with its radio transmitter attached.  The researchers captured 14 darners in New Jersey between September and October, glued the tiny bug-sized transmitters onto their thoraxes, and released them.  Then they tracked the dragonflies with radio receivers either by car or by Cessna plane for up to the 10 days of the transmitter’s life.  In essence, this qualifies as one of the most awesome research projects ever!  (Pardon me while I drool thinking about how amazing it would have been to track dragonflies from a plane in this study…)

Using this design, the researchers determined how far a dragonfly flew on any given day, how long it rested between flights, and the exact path it took during its migration.  They then put all of their data together to determine how similar dragonfly migrations are to bird migrations and what rules dragonflies follow when making migratory decisions.

So what, then, did they learn?  First, the dragonflies all migrated within within 4 days of receiving their radio transmitter, so they were still inclined to migrate even with the transmitter in place.  They also learned that the dragonflies tended to move approximately once every three days.  This means that the dragonflies flew one day, rested for two days, then flew again.  Long stopovers were apparently necessary during the migrations.

What about the distance and direction traveled?  The team found that there were three types of daily movements.  Some dragonflies flew a short distance (1-4 km) and in all sorts of different directions.  Others flew 8-12 km in a single direction.  Still others flew 25-150 km (that’s just over 93 miles) in one day!  Clearly these dragonflies were capable of flying long distances under certain conditions, though the average daily flight distance when all flights were combined was only 58 km (36 miles).  As for the direction, some dragonflies flew west and some flew east at times, but the bulk of the movement was southwest.

Weather seemed to be important for determining when the dragonflies flew and when they did not.  They were much more likely to fly in mild winds than in stronger winds, and no dragonflies flew if the wind speed was greater than 25 km/h (that’s just over 15 mph).  They also tended to fly more on days when the wind was blowing from the north than on days when the wind blew in other directions.  The dragonflies apparently depended on the wind to help them travel because the direction of the dragonflies and the direction of the wind on days where they flew were nearly identical.  Curiously, there was no association of temperature and propensity to migrate on flight days: there was no difference in the daily high temperatures of flying days versus non-flying days.  However, all migratory flights took place after a night with a temperature cooler than the previous night.

These data suggest that dragonflies have a set of simple rules they follow when deciding whether to migrate or not.  The dragonflies move with the winds (but not in very strong winds) in response to cool night and take a few days off between flying days, presumably to hunt and/or rest.  This in and of itself is pretty interesting, but it’s also interesting to place this information in the larger context of flying animal migrations.  Nearly everyone is familiar with the annual migrations of birds and know that birds “fly south for the winter.”  The data the dragonfly team collected revealed that the migratory patterns of dragonflies are remarkably similar to those of birds.  Songbirds use the same sorts of weather cues to prompt their migrations, follow coastlines and other prominent landscape features in the direction of the wind, and make frequent stopovers, just like the dragonflies did.  In essence, birds share the same set of rules governing migration that dragonflies exhibit.  It is likely that other migratory flying animals follow the same rules.

The team finished their discussion of the dragonfly behavior by using their data to calculate the maximum migration distance these dragonflies might be expected to fly.  Assuming a modest flight speed and a two month migration season, an individual dragonfly could be expected to fly 700 km, or 435 miles!  This is a long way for what is ultimately a small animal to fly.  Unfortunately, due to the limitations of the transmitters used (i.e. the battery life of 10 days), the team was never able to figure out exactly where the dragonflies ended up.  If the dragonflies are traveling 435 miles, I’ve calculated that dragonflies starting off in New Jersey most likely end up in West Virginia or Virginia.  This is much further north than previously suspected, which leads to at least two possible explanations for sightings of mass migratory swarms reported further south.  1) The dragonflies might fly faster than estimated, which would allow them to travel further during the 2 month migration season.  Or 2) the dragonflies observed in locations such as Florida and further south might be starting off from a more southern location to begin with.  Yet two more questions to be answered about this behavior!  It may be possible to answer these questions using the techniques the dragonfly team developed.  I suspect radio transmitters will play a significant role in answering some of the many outstanding questions about migratory behaviors in dragonflies.

Next time I’m going to post images and descriptions of the most common migratory dragonfly species so that people observing dragonfly swarms can determine which species they’re seeing.  In the meantime, I hope all your easterners enjoy watching the dragonflies that are on the move in your part of the country!  Based on the dragonfly activity in the north this year, it could be downright spectacular.

Literature Cited:

Wikelski M, Moskowitz D, Adelman JS, Cochran J, Wilcove DS, & May ML (2006). Simple rules guide dragonfly migration. Biology letters, 2 (3), 325-9 PMID: 17148394


Have you seen a dragonfly swarm?

I am tracking swarms so I can learn more about this interesting behavior.  If you see one, I’d love to hear from you!  Please visit my Report a Dragonfly Swarm page to fill out the official report form.  It only takes a few minutes!



Want more information?

Visit my dragonfly swarm information page for my entire collection of posts about dragonfly swarms!


Unless otherwise stated, all text, images, and video are copyright © 2010

From the Literature: Mass Migrations in Dragonflies

dragonfly swarm

Dragonfly swarming behavior. Photo copyright Steven Young and taken from

Wow!  I’m running way behind on getting a new post up.  I’ve been getting a ton of dragonfly swarm reports in the last week, so I’ve been scrambling around trying to keep up with that.  Couple that with some writer’s block and the start of my field season, and Ms. Dragonfly’s life has been a bit hectic!  I’ll try to get something up a bit more quickly next time…

I’m going to do a short series on dragonfly swarming behavior since this seems to be a topic that a lot of people are interested in this summer – and one that I’ve become rather obsessed with over the last few months.  (Seriously – ask my husband, my Ph.D. advisor, or my coworkers how many times I’ve brought up dragonfly swarms in conversation over the last few weeks!)  After combing through the literature, I’ve discovered some great information on dragonfly swarming and I think I have some good explanations for why we see these swarms in nature.  Today’s post will cover a paper that came out in 1998 about the mass migratory swarms observed in dragonflies.  For those of you out there who have seen hundreds of thousands or millions of dragonflies all flying together in a single direction, a sort of river of dragonflies, this paper describes and attempts to explain this behavior.  And what an amazing behavior it is!  I would dearly love to see one of these giant migratory swarms.  If any of you happen to get this behavior on video, I strongly encourage you to upload it to YouTube so that others and I can see it!  But let’s get to that paper.

Over the years, several dragonfly researchers have witnessed massive swarms of dragonflies flying together in what appeared to be migratory swarms.  Each time, they consisted of several hundreds of thousands or millions of dragonflies, enough to nearly blacken the sky and make themselves really obvious – if you happened to be looking up at the right time to see them!  In 1998, Robert Russell, Michael May, Kenneth Soltesz, and John  Fitzpatrick combined their eyewitness accounts of these migratory swarms and wrote one of the first papers describing this behavior in detail.  At the time of the paper’s release, very little was known about dragonfly migratory patterns.  25-50 species of all dragonflies were thought to be migratory and only 18 migratory species had been reported in the eastern United States.  In their paper, the authors set out to describe the massive swarm migrations they’d observed in dragonflies, but they also wanted to review the literature concerning migrations of dragonflies to see if any patterns emerged.  They were hoping to determine why these swarms formed and understand how they worked.

Anax junius adult

Green Darner (Anax junius) male – notice the large wings and huge eyes!

First the authors described three swarm migrations in the United States that three of the four authors had witnessed themselves: one swarm each in Chicago, Illinois; Cape May, New Jersey; and Crescent Beach, Florida.  In each place, the observers reported the same things:

  1. Nearly all of the dragonflies in each swarm were moving together in the same direction, though not all swarms moved the same direction.
  2. The swarms were largely made up of green darners (Anax junius – pictured to the left above), though several other species were present in the Cape May swarm to a much lesser extent (black saddlebags, twelve-spotted skimmers, swamp darners, Carolina saddlebags, wandering gliders, spot-winged gliders, and blue dashers).
  3. The swarms were very large, at least 200,000 dragonflies (Florida) and up to an estimated 1.2 million in the Chicago swarm!  (They compared the dragonfly swarms they observed to those observed in locusts and found that the number of swarming dragonflies was comparable to swarming locusts.  If you’ve heard about locust plagues, you have a good idea of what one of these swarms should look like!)
  4. The swarms tended to be very compact, that is they flew within a rather confined corridor rather than spreading far apart either horizontally or vertically.
  5. The swarms tended to follow very obvious landmarks such as shorelines, ridges, and coasts.
  6. All of the swarms took place soon after a cold front passed through the area, so they appeared to be distinctly weather-related.

The authors also reviewed the literature and collected field observations from other researchers to better determine how, where, and when these massive swarms were taking place.  They gathered 41 reports spanning the years 1881 through 1995 and found that these swarms generally shared the same characteristics as the swarms witnessed by the authors.  They further discovered that the swarms took place between July 30 and October 13, though the majority (28 swarms) occurred in September.  This was clearly a highly seasonal event.  All of these reports were also made in locations in the eastern United States, so it appeared that these swarms only occurred in half of the U.S.

Based on all of the reports together, the authors made some speculations about the massive migratory swarming behavior observed in several species of dragonflies.  First, their data suggest that at least some of the dragonfly species are migratory and are using these swarms to travel from north to south during the fall.  Where exactly they’re going is much less certain, but there are some reports of dragonflies crossing the Gulf of Mexico and massive swarms suddenly appearing in locations in Mexico.  The final destination could be as far south as Central America or as far north as the Gulf coast.  Regardless of where they end up, the dragonflies appear to be flying south for the winter, much like birds and monarchs do.

Second, the authors suggest that this type of swarming behavior might be unique to the United States and the regions where the dragonflies overwinter.  There are similar swarms in Europe, but the timing and recurrence of the swarms are very different than the yearly migrations observed in the US.  This also led the authors to suggest that the wandering glider, the only dragonfly species naturally found on all continents, might not form these migratory swarms.  While they have been observed in American migratory swarms, they typically only make up a small percentage of the total individuals within the swarm and are very uncommon in swarms reported in other locations.

Third, the authors proposed that weather may be the most important factor regulating the mass migrations of dragonflies.  Indeed, in nearly all of the accounts they collected, a cold front had passed through the area just prior to the appearance of the swarm.  A combination of season (autumn), north-westerly winds, and cold fronts may be necessary to entice dragonflies to fly from the cooler northern regions to warmer southern climates for the winter.  The timing of this behavior also suggests that these dragonflies are relying on aerial plankton for food as they migrate.  Aerial plankton is abundant during the fall, allowing the dragonflies to feed along the way.

Fourth, landmarks appear to be very important in directing dragonfly migration paths.  Mass migrations are almost always reported along major waterway, coasts, or prominent land features.  These “leading lines” help dragonflies orient themselves and follow the correct path south.  However, they also seem to reorient themselves under certain conditions, especially when they have to cross significant stretches of water.  This will be the subject of an upcoming post, so I won’t go into more detail here.

And last, there is some evidence that the dragonflies that fly south for the winter are not the same individuals that return north.  Dragonfly adults typically have a life span far shorter than the length of time the dragonflies overwinter, so it is very likely that migratory dragonflies fly south, mate, produce young, and die.  The offspring then make the return trip north in the spring.  There is a lot of uncertainty about what happens in the spring, however.  The route the dragonflies take on the return trip remains mysterious.  The migration north also does not appear to involve huge migratory swarms and may be comparatively inconspicuous, making it difficult to determine when and where the northward bound dragonflies are flying.

The authors ended their paper with a long list of questions that remain unknown about dragonfly migration.  12 years later, we still don’t know most of the answers to these questions!  For example, why exactly are these dragonflies forming these massive swarms when they migrate?  If it were only to protect the dragonflies as they fly (safety in numbers), then why do so few dragonfly species exhibit this behavior?  It doesn’t appear to be solely related to timing of adult emergence from the water either as emergence is not synchronous in the green darners and they are the most commonly observed migratory species.  Still, this paper went a long way toward explaining mass migrations in dragonflies.  Due to the ephemeral nature of swarms – a million dragonflies might fly over an area in a matter of minutes – we’re lucky there have been any studies of this behavior at all!

Over the next few posts, I’ll go over another paper about dragonfly migrations (one using radio transmitters!) and talk about what I’ve discovered in the literature about what I’m calling static swarms.  With these three posts, I hope to summarize most of what’s known about dragonfly swarming behaviors so that the information is available outside of the scientific community in an accessible way.  I am also going to post photos of all of the known migratory dragonfly species in the US.  Many people reporting dragonfly swarms to me have expressed an interest in identifying the dragonflies in their swarm, so I want to make that possible.  And, as always, keep sending me dragonfly swarm reports!  I have a huge collection of them going this summer and I couldn’t be happier with the level of participation in my project.

Literature Cited:

Russell, R., May, M., Soltesz, K., & Fitzpatrick, J. (1998). Massive Swarm Migrations of Dragonflies (Odonata) in Eastern North America The American Midland Naturalist, 140 (2), 325-342 DOI: 10.1674/0003-0031(1998)140[0325:MSMODO]2.0.CO;2


Have you seen a dragonfly swarm?

I am tracking swarms so I can learn more about this interesting behavior.  If you see one, I’d love to hear from you!  Please visit my Report a Dragonfly Swarm page to fill out the official report form.  It only takes a few minutes!



Want more information?

Visit my dragonfly swarm information page for my entire collection of posts about dragonfly swarms!


Unless otherwise stated, all text, images, and video are copyright © 2010

From the Literature: Beetles that breathe through their shells!

I wrote my last three posts on insect respiration specifically because I wanted to talk about a paper that was published in the Journal of Morphology in November 2009 that deals with aquatic insect respiration.  Since this is my personal area of expertise, I find this paper fascinating!  I hope you find it as interesting as I did.

One of the most common insects you’ll see in many freshwater systems is the predaceous diving beetle (Order: Coleoptera, Family: Dytiscidae).  As the name implies, these are beetles that live in water (hence “diving”) and they are fierce predators (hence “predaceous”).  They’re pretty amazing insects in many different ways and many of them are absolutely gorgeous to boot.  However, today I wanted to focus on one really interesting respiratory structure that a pair of researchers in Germany discovered.

Predacious diving beetles are generally atmospheric breathers and must go to the surface periodically to collect oxygen (see my posts on aquatic insect respiration and breathing more efficiently in water for more information).  They then carry air under their elytra while underwater.  Several species expose part of their air bubble to the water to allow them to take advantage of physical gill respiration as they swim around their environment.  However, some predaceous diving beetles are known to stay underwater a much longer time than would be expected if they relied solely on a physical gill.  Some of these beetles have been observed to have little air bubbles attached to their elytra, prompting some researchers to suggest that they might actually be breathing through their shells.

German researchers Siegfried Kehl and Konrad Dettner of the University of Beyreuth were interested in this observation.  They wondered whether certain species of predaceous diving beetles were able to use their elytra as a respiratory organ and set up a simple study to test the idea.  The team was interested in two things.  First, they wanted to know whether the elytra of the predaceous diving beetle Deronectes aubei were able to take up oxygen, that is whether they contributed to the respiration of these insects.  If so, they wanted to examine the structures on the surface of the elytra to determine which structures were contributing to respiration and how oxygen would be transferred from these structures to the rest of the body.

Meet Deronectes aubei:

Deronectes aubei

Deronectes aubei. Photo copyright Christoph Benisch and taken from

This is one of the smaller predaceous diving beetles and is just over 1/8″ in length.  This particular species is European and found in Germany, so Kehl and Dettner were able to observe them in their lab.  They discovered that this species is capable of remaining submerged without any access to atmospheric oxygen for over six weeks!  Because these beetles live at the bottom of very fast flowing water, anything they can do to increase their submersion time will help prevent their bring washed downstream and/or damage to their bodies.  Being able to breathe through their elytra and absorb more oxygen from the water would be an excellent adaptation for living in this environment.

First Kehl and Dettner determined whether D. aubei was able to take up oxygen via their elytra using a technique called respirometry.  Respirometry is rather complicated to do and there is a ton of math involved, but the concept is simple.  First, you measure the oxygen content of air or water entering a chamber where an organism is housed, in this case a beetle or group of beetles.  Then you measure the oxygen content of the air or water after it has been exposed to the organism for some length of time.  The difference between the two measurements is the amount of oxygen that the organism used.

Kehl and Dettner measured the oxygen uptake of the beetles from the water using two different types of respirometry.  They also measured the oxygen uptake of beetles in two different treatments.  First they painted a resin onto their elytra, effectively blocking them from taking in oxygen.  Then they removed the resin and measured them again.  To determine whether the beetles were taking up oxygen via their elytra, they compared the oxygen uptake of beetles with the resin covering their elytra (the treated beetles) to the same individuals after the resin was removed (the control beetles).

Both types of respirometry provided the same results: beetles exhibited a significantly larger oxygen uptake when their elytra were exposed to the water than when the elytra were coated with resin and not allowed contact with the water.  These results mean that the beetles are taking up oxygen via the elytra and are using them as a respiratory structure.  The team also found that blocking the physical gill (that air bubble they expose to the water) did not significantly decrease the oxygen consumption of the beetles.  This told them that the beetles are getting most of their air through their elytra and that the air bubble they carry isn’t that important.

scanning electron microscope

The scanning electron microscope I use for my work. The sample goes into the chamber on the left and the image pops up on the computer screen on the right. The electron gun is the cylindrical part sticking up above the sample chamber.

Kehl and Dettner conclusively showed that the beetles were capable of taking oxygen in via their elytra, so the next step was to determine how exactly the elytra are used as a respiratory organ.  To do this, the team examined the structure of the elytra using electron microscopy.  Electron microscopy, like respirometry,  is rather complicated, but for now just know that there are two different types.  Scanning electron microscopy (SEM) is used to look at the surface features of an object and the samples can be pretty big, as large as the sample chamber will allow.  An electron gun is aimed at the object, they bounce off the sample, and sensors collect them to create an image of the object on a screen.  In contrast, transmission electron microscopy (TEM) is usually used to look at the internal structures of an organism, but you can only look at very small portions of the object of interest at a time.  An electron gun is aimed at a very, very thin slice of the thing to be examined and a sensor below the sample collects the electrons that pass through the sample to create an image on a screen.

Kehl and Dettner used both types of electron microscopy in their study.  First they used SEM to look at the surface structure of the elytra of D. aubei.  They discovered that there are three different types of hairs present on the beetles.  One is a simple hair that is found in deep pits.  Another is a hair that is found in shallow pits.  The third is a spoon-shaped hair with an enlarged base that is also found in pits.  Hairs are used for a lot of different things in insects including sensation (smelling and/or feeling the environment around them) and proprioception (determining which body parts are positioned where), so the team used TEM to see which specific hairs were involved in respiration.  They found that the spoon-shaped hairs contained tracheoles that traveled down cylindrical channels in the elytra.  These tracheoles then connected to one of four big tracheae inside the elytra that carried oxygen to other parts of the insect.  This meant that the spoon-shaped hairs were the respiratory structures of the elytra, the areas where oxygen was being taken into the elytra to be used by the insect.

What all of this means is that these beetles have tiny gills littering their elytra that they use to absorb oxygen from the water!  Kehl and Dettner call them setal tracheal gills.  These gills likely significantly increase the surface area through which the beetles can absorb oxygen, allowing them to remain underwater much longer than they could without them.  Because the beetles live in fast flowing, cold water, they live in an ideal aquatic oxygen environment.  These gills allow the beetles to make the most of the abundant oxygen.  The beetles also live in a high risk environment for moving around, so the longer they can stay underwater the better.  Kehl and Dettner suggest that other beetles that either live in very fast flowing waters or that live far under the substrate in streams likely have similar structures as well.  It is common for organisms that live in similar habitats to exhibit similar structural and/or behavioral features that suit the environment, even if they are widely disparate species.  The team predicts that many beetles living under gravel in fast flowing streams might have the same respiratory structures as D. aubei.

I believe that this study is a particularly good one.  Aside from my personal interest in the subject, Kehl and Dettner simply did a great job in creating an elegant study.  First they determined conclusively that the beetles were able to use their elytra to take up oxygen by comparing the oxygen consumption of beetles that were allowed to use their elytral structures to the same beetles when they could not use their elytra for respiration.  Then they used SEM to identify potential respiratory structurs on the elytra.  Finally, they used TEM to determine how the structures on the surface were connected with the internal respiratory system of the beetles, identifying the specific structures involved in respiration and the pathway oxygen takes to get into the beetle’s respiratory tract.  They showed both that the beetles were using their elytra as a respiratory structure and how they were able to do so.  Then they finished off their paper by suggesting the conditions in which this sort of respiration might be particularly well suited and made predictions that could extend to several other species.  It’s an example of a very neat, tidy experiment that uses simple experiments to discover new things with broad implications, which is what ever researcher hopes to achieve with his or her writing.

This concludes my series on insect respiration and aquatic respiration in insects!  If you have any questions or would like more information, please feel free to leave a comment.  Until next time!

Literature Cited:

Kehl S, & Dettner K (2009). Surviving submerged–Setal tracheal gills for gas exchange in adult rheophilic diving beetles. Journal of morphology, 270 (11), 1348-55 PMID: 19480011


Unless otherwise noted, text and images copyright © 2010

From the Literature: Jocks and nerds in the damselfly world (The Dragonfly Trilogy, Part Three)

Welcome to the third and final segment of the Dragonfly Trilogy – and another installment of From the Literature!  If you don’t know anything about dragonfly territoriality, I recommend reading part two of my trilogy for more information on how dragonflies and damselflies set up and defend territories.  You’ll get more out of this post if you know something about territoriality before diving in.

Last time, I discussed how odonates benefit from being territorial, why they set up territories and defend them from other males.  Like many things in biology, it boils down to sex: males that defend high quality territories generally mate with more females than males with low quality territories.  Likewise, males that defend territories generally mate with more females than males that do not defend territories.  It is usually better to be a male with a territory than a male with no territory, but there are often many more males than there are available territories and some males are inevitably left out.  Presumably the stronger, better males (the most fit males) end up successfully claiming and holding territories while the weaker, wimpier males are left without territories and become wanderers.

Calopteryx virgo male

Calopteryx virgo male. Photo from Wikipedia, by Michael Apel.

This is the situation that a group of researchers  in Finland recently investigated.  They chose to study the damselfly Calopteryx virgo, a European damselfly also known as the beautiful demoiselle.  This gorgeous creature is pictured at left and is one of the damselflies known to be territorial.

The researchers asked a simple question: are C. virgo males that defend territories larger than males that do not defend territories?  They wanted to know if the damselflies that were able to protect a territory from other males were somehow better suited to being territorial than the damselflies doomed to be wanderers and unable to gain their own territory.  They also wanted to know if this changed over time, whether the non-territorial males eventually became territorial.

To answer these questions, the researchers captured mature C. virgo males arriving at their study stream in Finland, then marked them (so they could tell them apart), measured their right hindwings, weighed them, and released them back into the study area.  They observed the damselflies on two different days ten days apart and determined whether the males were territorial (they stayed within a small, 2 meter area for at least three hours) or non-territorial (they did not remain within a 2 meter area).  Males were considered wanderers if they moved more than 100 meters during the observational periods.

The team discovered that there was a significant difference in size between territorial and non-territorial males.  Territorial males had longer wings and were heavier than wandering males, so the bigger males were the ones claiming and holding territories.   The researchers also discovered that time didn’t have much to do with whether a damselfly male was territorial or not.  The wing length and weight of the wandering males was about the same and wandering males were consistently smaller than the territorial males on both days.  Wandering males made up about the same percentage of the population both days too.  So, the smaller males weren’t ever getting territories and were consistently being excluded.

The data that Koskimaki and colleagues presented in their paper suggest that the bigger, more physically impressive males get more mates.  To better understand the significance of these results, let’s consider a similar situation in humans that many people will recognize.  Think about what you know about stereotypical high schoolers.  Who gets more dates: the jocks (usually the bigger, more physically impressive males) or the nerds (usually the smaller, less physically impressive males)?  When I was in high school (I myself was firmly rooted in the nerd category!), the jocks got most of the girls while the nerds usually only admired the girls from afar.  The jocks outcompeted the nerds physically, and because they were generally more attractive, they excluded the nerds from finding dates by hogging all of the available girls.  If you ignore the possible confounding affects of wealth, intelligence, and overall personality that come into play in human mating behaviors, almost the same thing is happening in the high school students that we see in the damselflies that Koskimaki and colleagues studied!  In effect, the jocks among the damselflies were getting all the girls because they were better suited to protecting territories, and thereby attracted more mates, than the nerds who were unable to gain a territory.

I’ll end with two important questions: if it is so much better for males to be bigger so that they can more successfully hold valuable territories, 1) why are any males small and 2) why aren’t the damselflies getting bigger and bigger over time?  Koskimaki and colleagues suggest that that territorial and non-territorial males might form two distinct subgroups within the damselflies, each with their own strategies and goals for mating.  Even males without territories are able to mate with some females.  They could end up with the same number of offspring, thus ensuring the continued existence of smaller males in the population, if they have means for compensating for their relative lack of mating opportunities.  The team cites several other studies that suggest that this is happening in several territorial damselfly species, that non-territorial males are equally successful in producing offspring compared to territorial males.   It is likely that there are some benefits to being smaller or some costs to being larger that have not yet been accounted for.  Further studies in this area would be a great avenue for future research!

I hope you’ve enjoyed the dragonfly trilogy!  It’s been a lot of fun delving into the dragonfly literature for a few weeks and sharing information about my favorite group of insects.  I’m sure to post more odonate research in the future, but next time I’ll be telling a story of a centipede and a woman who is very, very scared of them – me!

Literature Cited:

Koskimaki, J., Rantala, M.J., & Suhonen, J. (2009). Wandering males are smaller than territorial males in the damselfly Calopteryx virgo (L.) (Zygoptera: Calopterygidae). Odonatoligica, 38 (2), 159-165.


Text copyright © 2010

From the Literature: Nice Guys Get the Girls!

I love aquatic insects!  They do some amazing things and are incredibly interesting animals.  That said, I feel like most people know very little about aquatic insects and the role they play in our world.  Heck – some people don’t know that aquatic insects even exist!  So, for my first From the Literature post, I thought I’d discuss a recent paper dealing with aquatic insects.

In biology, it is thought that males benefit from mating with as many females as possible.  Because males usually do not care for their offspring and invest little in producing sperm, it is best for them to mate with as many females as they can, thereby contributing their genes to as many offspring as possible.  There are, of course, exceptions to this general rule (the giant water bugs I study are an excellent example!), but it holds true for many species.  Because females usually make a greater investment in their children – if nothing else, eggs are much more nutritionally expensive to produce than sperm – they often cannot mate as often as males.  As they contribute their genes to fewer offspring, it is to their benefit to choose the best mates, the ones that will likely produce strong and robust children that have a high chance of surviving to adulthood.  In essence, there is a battle of the sexes going on: males want to mate all the time with every female they can find while females want to mate with only the best males.  It pays for a male to be aggressive and secure as many mates as possible while it pays for a female to be choosy.  In essence, there is a trade-off between what males want and what females want: when one sex succeeds, the other suffers.

water strider interactions during mating

In this image by Omar Eldakar, a hyper-aggressive male (on the right) attempts to break up a mating pair of water striders. The colored dots were used by the researchers to keep track of individuals.

This idea sets the stage for a recent paper by Dr. Omar Eldakar (currently a postdoctoral fellow at the University of Arizona) and his colleagues published in the November 6, 2009 issue of Science.  Eldakar and his team studied a type of aquatic insect called water striders (order: Hemiptera, family: Gerridae).  Water striders (also known as Jesus bugs, water spiders, and pond skaters) are long, skinny insects that live and hunt on the surface of water.  Water striders are typically found in groups of several individuals called aggregations in calm areas of streams and ponds.

The species Eldakar studied is Aquarius remigis and it is well-known for its battle of the sexes.  Many male A. remigis individuals are highly aggressive when pursuing females, lunging at and jumping on their potential mates.  While the females often resist  mating with these hyper-aggressive males, the behavior has been known to improve mating success.  Aggressive males are usually more successful at securing mates than non-aggressive males.  This leads to a question: if aggressive males mate with more females than non-aggressive males, why aren’t ALL males aggressive?  Eldakar and his colleagues wondered if the fact that most studies of sexual conflict in water striders do not allow individuals to migrate between aggregations might explain why hyper-aggressive males are reportedly so successful.  If females are forced to remain in an area with hyper-aggressive males, the males might have a higher mating success than if the females could be choosier about who they mated with.

Eldakar and his colleagues set up an ingenious experiment to test this idea.  First, they placed water striders in an artificial pond and observed male aggression, movement of females, and mating attempts/successes.  Water striders were able to form their own aggregations and move freely between them.  The researchers then divided the pond into several sections, placing males of various aggression levels with females in each section.  The same observations were made, though this time individuals were not able to move between groups.  Finally, the group compared their observations of the open treatments to the closed treatment to see how movement contributed to the mating success of aggressive and non-aggressive males.  They discovered some interesting things.

As had been reported in other water strider studies, hyper-aggressive males had more successful mating attempts than the non-aggressive males in the closed system.  If females were not able to move to another aggregation (i.e. their choices in mates were restricted), they mated more frequently with the aggressive males than the non-aggressive males.  However, when the striders were able to move between groups, Eldakar observed that females moved to other aggregations when harassed by aggressive males.  Aggressive males repelled the females they wanted to mate with!  The females would, however, readily mate with the less aggressive males in their new areas.  This meant that less aggressive males were able to secure many more mates in open systems than in closed systems.  In other words, the females avoided the aggressive jerks and the nice guys were getting the girls!

I think this is a great paper.  The experiment was very simple, but it ended up revealing a lot of information.  Many scientists call this sort of experiment “elegant” and I think this research certainly qualifies as an elegant experiment.  Eldakar was able to refute the findings of several previous papers with an experiment that should be very easy for others to duplicate, one of the conditions for good science.  He also clearly reminded biologists of one of the perils of doing behavioral research in the lab: unnatural conditions sometimes lead to unnatural behaviors.  As a behavioral researcher myself, I think this isn’t reinforced enough.  By modifying what had been done in the past, by allowing aggregations of water striders to form naturally rather than forcing them into pre-determined groups, Eldakar learned that aggressive males are not always the most successful in nature.  This corrected what we already knew and gave us some great information that will be useful for many future studies.  Thanks, Dr. Eldakar!

Literature Cited:

Eldakar OT, Dlugos MJ, Pepper JW, & Wilson DS (2009). Population structure mediates sexual conflict in water striders. Science (New York, N.Y.), 326 (5954) PMID: 19892974


Text copyright © 2009

This post was chosen as an Editor's Selection for This blog post was an Editor’s Choice selection at!