Schneier on Security
A blog covering security and security technology.
« Shower Mirror with Hidden Camera |
| Michael Froomkin on Identity Cards »
March 3, 2009
Three Security Anecdotes from the Insect World
Beet armyworm caterpillars react to the sound of a passing wasp by freezing in place, or even dropping off the plant. Unfortunately, armyworm intelligence isn't good enough to tell the difference between enemy aircraft (the wasps that prey on them) and harmless commercial flights (bees); they react the same way to either. So by producing nectar for bees, plants not only get pollinated, but also gain some protection against being eaten by caterpillars.
The small hive beetle lives by entering beehives to steal combs and honey. They home in on the hives by detecting the bees' own alarm pheromones. They also track in yeast that ferments the pollen and releases chemicals that spoof the alarm pheromones, attracting more beetles and more yeast. Eventually the bees abandon the hive, leaving their store of pollen and honey to the beetles and yeast.
Mountain alcon blue caterpillars get ants to feed them by spoofing a biometric: the sounds made by the queen ant.
Posted on March 3, 2009 at 1:20 PM
• 23 Comments
To receive these entries once a month by e-mail, sign up for the Crypto-Gram Newsletter.
Anyone else find themselves reading this in Tom Waits' voice?
I am now, Sean. You made my day!
Reminds me of the connection between the life cycle of cicadas and prime numbers.
"You also find prime numbers in the life cycles of cicadas. There are about 1,500 species of cicadas known. There are those that appear yearly in midsummer, and there are also the so-called "periodic" cicadas. They appear at prime number intervals - 7 years, 13 years and 17 years.
The cicadas are part of the insect order Homoptera. These are all sucking insects, which pierce plants with their pointy mouthparts and suck out the juices. The breeding cycle begins when huge numbers of adult cicadas emerge in the spring. They mate within a week, and a few days later, the female lays her eggs. She drills into the wood of trees, and inserts up to some 400-to-600 eggs. These eggs hatch up after two to six weeks. The little babies make their way down to the ground (by crawling down, or just dropping), dig their way into the soil with their claws and begin the next phase of their life, feeding on the roots of shrubs and trees for the next 6, 12 or 16 years. The 17-year cicadas are almost fully grown into nymphs by 8 years, but they continue to feed underground until the 17th year when they come out of the soil, and attach themselves to any nearby tree or post. Their shell splits open, the adults emerge and live only for a few weeks before dying.
Now biologists have asked for a long time whether it's just a coincidence that the emergence period of the three species of periodic cicadas (7, 13 and 17 years) are all prime numbers.
One previous theory was that if the cicadas are running on different cycles, and if these cycles are prime numbers, they'll cross over only very rarely. For example, a 13-year cycle and 17-year cycle will meet only every 221 years. That means that both species of cicadas would come out in huge numbers and all have to compete for the same amount of food only once every 221 years. The rest of the time, there would be enough food.
This is a nice theory, but Mario Markus, a physicist from the Max Planck Institute for Molecular Physiology in Germany has come up with a new theory. It's related to periodic predators. Suppose there are some predators (like birds, and the Cicada Killer Wasp) that attack cicadas, and that the cicadas emerge every 12 years. Then the predators that come out every two years will attack them, and so will the predators that come out every 3 years, 4 years and 6 years. But according Mario Markus, "if the cicadas mutate to 13-year cycles, they will survive."
If the periodic predator comes every 3 years, the best plan for the cicadia is to also come every 3 years (but in one of the 2 years that the predator never shows up). Instead of having predators around every third cycle, they'd *never* have those predators around. The best survival strategy is to have a regular frequency that is aligned with the predator, but to add an extra year whenever a large predator population is encountered - to get back to an unsynchronized cycle.
IANAEB* but I wonder if there is a greater chance of a predator on the same period/different synchronization as in John Macdonald's post stumbling onto a perfect feeding opportunity. Something delays the lifecycle one year, and they suddenly dig themselves up surrounded by food, and it continues every cycle the prey is extant.
(I am not an evolutionary biologist)
As far as the cicadas go, I'm sure that both factors are important to some degree, but I have to wonder how many "periodic predators" there are to begin with. I certainly don't know of any non-insects that appear less frequently than once a year (with the exception of those desert amphibians that hibernate indefinitely waiting for the rains to come so they can spawn, not exactly relevant in this case). Furthermore, as Sean points out, any predators whose lifecycle happened to drift into sync with the cicadas would have a huge advantage and soon outnumber their out-of-sync relatives - and there will always be a few differently-timed mutants in every generation, so this would happen fairly quickly. The non-competition hypothesis also has the advantage of explaining why different species have arrived at different prime numbers.
That helps you against the current predator, but not the next cyclical predator. Having a prime period would protect you against any predator that doesn't have a cycle that's a multiple of your prime number.
There are the obvious defence of butterflies and peacocks and other creatures with large "eye" paterns to scare of birds and other hungrey criters.
Then there are the sand and hover flys that look like wasps and other mimic insects.
My fav though it the "golden poisend arrow or dart frog" and it's cousins. These tiny frogs get their poison from the ants they eat and tend to live in or around bromiliads. They are so leathal that they need no camoflage to hide behind. Luckly when kept in captivity as they no longer eat the appropriate ant the very quickly lose their poison making ability.
Oh and the peacoks tail is an interesting example of evolutionary preasure that also could apply to security. The larger the tail the more likley a male is to find a mate, however as the tail gets larger it is less able to evade preditors...
With regards to prime numbers and nature they do have a habit of poping up all over the place, why is unknown but nature seems to like them.
Also, IANAEB. If you read Richard Dawkin's Extended Phenotype, there are a large number of examples on how adaptation works in evolutionary biology and there are a surprising number of parallels to security and dealing with risk. For example, the co-evolution associated with a Red Queen race looks like what is seen with malware/anti-malware evolution.
I like the natural world and its parallels to security and warfare are uncanny.
Check out this behavior from the Prying Mantis:
...When flying at night, at least some mantises are able to detect the echolocation sounds produced by bats, and when the frequency begins to increase rapidly, indicating an approaching bat, they will stop flying horizontally and begin a descending spiral toward the safety of the ground, often preceded by an aerial loop or spin...
@ Sean & Ben
Not only in Tom Waits voice (thanks for the suggestion) but my thought was the hive beetle was making mead from the honey with the yeast, thus earworming self with Waits' "The Piano Has Been Drinking..."
One also thinks of Rome falling to the Barbarians, Visigoths and Vandals, who then loot and pillage until it all runs out.
But hey, there's always another hive, right?
another example - lap-dogs
a veterinarian told me once they were bred and chosen in medieval times by the rich to attract fleas and ticks - the higher body temp of a dog and thinner skin seem like a better and easier food source
Many moths, which fly at night and are fed upon by bats using echolocation, will fold their wings up and drop like stones when a bat's feeding buzz(*) is directed at them, and much of the time that's enough for them to escape. Insects that do not do this are caught nearly all the time.
(*) the feeding buzz is when a bat is homing on something to catch it; rather than a couple of clicks a second as a general sweep of stuff in the area, they emit a very rapid series of clicks, increasing in frequency as they approach their in-flight meal.
@JRR: sounds similar to a weapons control radar in an aircraft; it sweeps all around, but when it locks on a bogey to direct a weapon there, it radically increases hits.
Concerning the bees, since the bees are less sensitive to the bee alarm pheremone than the beetles are, presumably we can set out traps around beehives that emit low levels of bee alarm pheremone (assuming such can be synthesized) and the beetles will be drawn there and can be neutralized without affecting the bees.
Re: Why these three?
Why, indeed? There is another prime number in the same range, 11, which seems curiously absent. Why is the nearby 13 selected for, but not 11?
Every one of these struck me as having not only a basic similarity to a security issue, but additional similarities in the response of the defenders.
1a. Beet armyworms are the good guys.
Your virus scanner automatically quarantines and deletes detected malware. Unfortunately, the last update started flagging your word processor as malware, and now everyone's word processor has been deleted.
1b. Beet armyworms are the bad guys.
You don't actually have to be at home to deter a risk-averse burglar. You can have a night-light that not only allows you to get up easily at night, but also gives the impression of being home.
2. Hive beetles are the bad guys.
They use an infrared detector to find infrared perimeter security systems. Then the interrupt them to trigger a security response, but disappear without entering. They repeat this until the owner of the property turns off the security system because it is proving useless; without proper security, the owner actually moves away. Then they move in en masse.
3. Mountain alcon blue caterpillars are the bad guys.
The use a recorded voice to gain entry to a voice-print controlled cafeteria. Since the cafeteria only has perimeter security, once the entry criterion has passed, they get fed just like the ants.
The beet armyworm caterpillar story is amusing. Among slightly higher lifeforms, I've always been amazed at how good shorebirds are at distinguishing different predators. If a fish-eating Osprey flies over, they continue feeding undisturbed. However if any other large raptor glides over (especially a falcon but hawks too), the entire flock is in the air almost instantaneously. Somehow they can tell the difference between a non-threatening Osprey and a very threatening Red-tailed Hawk or Peregrine Falcon.
Er...don't plants produce *nectar* for bees, not pollen? Yes, they produce pollen, but that's for reproduction. The nectar attracts the bees (who use it as a food source, turning it into honey) and just so happen to pick up enough pollen to fertilize other plants along the way.
Mackenzie: The bees also snag a fair bit of the pollen, turning it into "bee bread". (Hit Google, my links just got evaporated by preview). Remember that pollen is plentiful, and it only takes one speck to fertilize each seed.....
Nectar is a sugar solution and provides an energy source. To grow from egg to adult, bees need protein, and that comes from pollen.
Schneier.com is a personal website. Opinions expressed are not necessarily those of BT.