Great glowing spiders

Blogging on Peer-Reviewed Research

I've known that scorpions have fluorescent cuticles — if you go out into the desert with a black light and shine it on the ground, the scorpions will often glow green and blue and be easy to spot. I had no idea that many spiders exhibit the same phenomenon, but there they are, glowing away. I may have to visit my local head shop (in Morris? Hah!) and get some black light bulbs to see what the fauna in my living room is up to.

Fluorescence is actually a fairly common property: all it requires is a molecule called a fluorophore that can absorb and capture transiently photons of a particular wavelength, or energy, and release them at a lower energy. What this means is that a fluorescent substance absorbs light at one range of wavelengths, and then re-emits those photons at a longer wavelength; there is a color shift. In the case of black light posters and spiders and scorpions, they are absorbing light at wavelengths our eyes can't detect (wavelengths below about 400nm, or ultraviolet light) and shifting it to a wavelength we can see, for instance to a nice blue at 450nm, cyan at around 500nm, or green at about 550nm. So to test this, all you need is a dark room, a spider, and a light source that glows at the wavelength that is absorbed by the fluorophore, and a detector (like, say, your eyes) that can collect light at the emission wavelength.

Or you can use a fluorimeter, a device that captures and measures the intensity of light emitted at specific wavelengths. Here's what's measured when you take various spiders, shine a UV light at them, and record the intensity of the emission at various wavelengths. Those are some respectable Stokes shifts from the spiders.

i-9104c0c8adf4df79c2b0cd31dd690b65-spectra.gif
;Normalized emission spectra of fluorophores from four spider species. Araneus diadematus (Araneidae) spectra are coloured red (288 and 330nm excitation). Dysdera crocata (Dysderidae) spectra are coloured dark blue (290 and 328nm excitation). Enoplognatha ovata (Theridiidae) shows only a single fluorophore peak (coloured gold), with excitation at 291nm. Misumena vatia (Thomisidae) shows a bimodal spectrum (coloured light blue) with excitation at 333nm, consistent with the presence of two different fluorophores with similar maximal excitations.

I know, charts and graphs are so bloodless and boring. What would you see in this situation? The photos below are of two species of spider, photographed on the left with normal visible light illumination, and on the right with UV illumination. Oooh, they glow!

i-2731b2548ff094da57889c618d3090d0-fluorescent_spiders.jpg
(a) Micrathena gracilis under white light (i) and UV illumination (ii). Note that entire cuticle of the abdomen fluoresces under UV, despite dark
colouration of protuberances under white light. (b) Hyptiotes sp. under white light (i) and UV illumination (ii). Note that white setae on anterior abdomen fluoresce, whereas dark setae
on posterior abdomen do not. (c) Distribution of visible fluorescence across spider families. Each square corresponds to one species sampled and colours
denote intensity of fluorescence.

Notice also the taxonomic distribution: it's all over the place. All of the spiders have fluorophores in their blood — if you do an ethanol extraction of the hemolymph, you can get nice glowing test tubes of fluorescent spider blood — but they vary in how much of the fluorophore (whatever it is; it's unidentified here) they deposit into the cuticle and into more thinly shielded tissues, like joints and hairs, that are exposed to the light.

I do disagree with the authors on one thing, though. They try to argue that this variation implies evolutionary significance, and I don't think it does. This kind of spotty distribution that doesn't show much coherence in the lineages implies to me that it is simply random variation in a trait that doesn't have much of a selective advantage or disadvantage. Fluorophores are ubiquitous, and I wouldn't be at all surprised if the spider fluorophore was simply an incidental property of something like spider hemocyanin, and its deposition pattern is a neutral phenomenon.

Now maybe it could be important. Fluorophores can act to enhance color in visible wavelengths (detergent manufacturers add fluorescent chemicals to laundry soap, for instance, to make "whites whiter than white") and if these spider species were using color cues for courtship or camouflage, then maybe a case could be made. The authors do point out that flowers reflect light in UV wavelengths and some insects recognize and target flowers on the basis of colors we can't see, so this could be important for camouflage, either to blend in or perhaps, by absorbing and thereby muting light at wavelengths that prey are tuned in to recognize, they are hiding themselves. The paper really doesn't do any experiments to test their evolutionary hypotheses, but does show that spiders have some interesting color properties to which we humans with our limited eyes may not pay much attention. Unless, of course, we start decorating our homes in the styles of the 1960s.


Andrews K, Reed SM, Masta SE (2007) Spiders fluoresce variably across many taxa. Biology Letters 3(3):265-267.

More like this

Don't have to go to a "head" shop. Just drop in a pet store and get a battery-powered blacklight used to locate cat urine on the carpet.

Cheap and portable.

No need for the long extension cords in the desert.

Gary

My favorite illustration of the fluorescent properties of scorpions was in a nature documentary a few years ago. Some researchers went into a tiny home in rural Mexico and offered to locate all the scorpions present for the owner. When the lights were switched off and the blacklight switched on, pretty much the entire room started to glow. Ick.

Edmunds sells a cheap ($5) keychain-size UV LED flashlight that works well.

Woohoo. Psychadelic spiders, too? In the 70's, I lived in the country when I was a kid and the black light in my bedroom illuminated scorpions that had infiltrated the house every so often. Very freaky and no weed was involved...

Portable blacklights (and replacement tubes) are available wherever rockhounds shop. Fluorescence is helpful in finding and identifying some minerals.

By Ken Shabby (not verified) on 12 Dec 2007 #permalink

Spiders... now even more awesome than I once suspected. I grew to love the little things in college. The dorm I stayed in was an old hotel, so the rooms opened to the outside of the building. Because of all the lights over the doors, it was a flying insect magnet, and as such was a spider magnet, and the spiders did their part to ensure that we rarely had moths, flies, or wasps in our rooms. I'm guessing years of such an environment had shaped the general spider behavior because the rarely built their webs across the walkways or doors, they just built around the lights with some pretty impressive webs. I now have the urge to drive back out there with a black light on a night when the power is out to see all the little glowing spidies.

By HumanisticJones (not verified) on 12 Dec 2007 #permalink

A spandrel? Likely. Still it seems to me that the branch containing the Dysderidae down to Uloboridae in the figure above are enriched for "intermediate" to "bright" while the group that ranges from Agelenidae to Lycosidae are enriched for "intermediate" to "dim".

By shiftlessbum (not verified) on 12 Dec 2007 #permalink

Wow! I did not know spiders were fluorescent. (My GF is crippled by arachnophobia, but she also thinks my UV strobe light is silly. Maybe she'll give me a break if I find a spider for her.

Not that anyone cares but I bought some UV LED's from Mouser a couple months ago. (Seoul semiconductor PN: LU531) and hooked them up to a 555 timer chip.)

Very interesting - it immediately gave me the idea that this property could be used when surveying spiders (which I do occasionally). All survey methods are biased, of course, and this is no exception. But it has potential for locating rare species.

Thomisidae you would expect to mimic the flourescent properties of flowers, since most of them hunt in a sit-and-wait fashion on flowers. I can see how flourescence could be a disadvantage in some groups - including Araneidae (highly flourescent in the graphic), which are orb-weavers and would therefore not 'want' to advertise their presence to passing flies who are tuned in to the flourescent properties of flowers. That's a little odd, since Araneidae are often to be found sitting in the hub of their web - why? It must be quicker to move to caught prey from there, which might otherwise escape - but presumably this has to be balanced against being more visible. It's possible their behaviour is affected by lighting, but I don't know if any work has been done on this.

I learned about fluorescence in clothing years ago (not sure if it was from detergent, or the dyes used) in Pirates of the Caribbean at Disneyworld. I was wearing a red blouse with white trim, and in the middle of the ride I looked down, and the white trim was glowing. Gave a great hint about how the effects were generated.

Assuming PZ's hypothesis is true, and the variation is not due to selection, what would be the proper term for that? Nonaptation? Spandrel? What term is most commonly used to denote this sort of thing?

Re #8

Yeah, I agree. There does appear to be a slight enrichment between the two main clades.

Be careful, though. Dyseridae and Pholcidae are basal on the tree, not part of that Theridiidae-Uloboridae clade (remember, on cladograms, branches can spin around their nodes all they want, its the branching pattern that is important).

There is likely a sampling issue to watch out for as well. Each blue box represents a sampled species within that terminal taxon. The "enriched bright" taxa appear to generally have more sampled species than the "enriched dim" taxa.

With this distribution, I'd be tentative about making any claims of differential expression across the tree. More data might elucidate a pattern, if there is any, though.

By Prillotashekta (not verified) on 12 Dec 2007 #permalink

This same technique works for finding Elvis in the desert.

By fardels bear (not verified) on 12 Dec 2007 #permalink

Ahem. In my previous message (@ #13), that first sentence should be modified with the phrase:

"...'appear" being the operative word"

We now return you to your regularly-scheduled comments thread, already in progress.

By Prillotashekta (not verified) on 12 Dec 2007 #permalink

Yup, UV LEDs are pretty easy to find (white LEDs are just UV LEDs in a flourescent package, iirc). Somewhere I have a pen with flourescent but otherwise transparent ink that's got a UV LED for reading by.

I've heard birds can see UV; maybe these chemicals make spiders harder for birds to see and catch. Though I suppose a guess at a function would have to explain why there's such wide variation.

Hmmm. I'm not sure what my arachnophobic friends will think about this - on the one hand they might be better able to 'purge' their surroundings of spiders (not that this is any kind of noble goal...) but on the other hand they might now find that there were a disagreeable number of spiders around that they had previously been blissfully unaware of.

Personally, I think this adds to the coolness of spiders (which I have to admit to not being entirely comfortable around) and am going to try and look for glowing spiders just as soon as I can get a blacklight.

This same technique works for finding Elvis in the desert.

How many times have you found Elvis using it? Or have you found more than one Elvis? :-D

The authors do point out that flowers reflect light in UV wavelengths and some insects recognize and target flowers on the basis of colors we can't see, so this could be important for camouflage, either to blend in or perhaps, by absorbing and thereby muting light at wavelengths that prey are tuned in to recognize, they are hiding themselves.

I've heard birds can see UV; maybe these chemicals make spiders harder for birds to see and catch.

I assume both predators and prey can also see "visible" light, so this doesn't seem so useful.

This awesome post reminded me that I still haven't finished reading The Ancestors Tale. I've gotsta get back on that!

By Robert Maynard (not verified) on 12 Dec 2007 #permalink

very interesting, thanks!

did you forget the "blogging about scientific research" icon, or have you decided you're not bothering with that any more?

my local electronics parts shop (used to be Marvac Dow, I think it's now DuVac or DoVac or something) sells a 6 or 8 LED UV flashlight that's pretty bright for about $13. It also has a little visible light, which is a bit less black-light-ish but does let you know where it's pointed. It seems to have a better frequency range for fluorescent inks than a headshop-style blacklight bulb I got.

Doesn't human skin pigmentation show the same pattern of spotty distribution across lineages? Tropical peoples have dark skin, and high-latitude peoples have light skin, regardless of lineage. Clearly that's adaptive.

Same goes for stripes among equids. (See Gould's "What, If Anything, Is a Zebra?") Species that need camouflage have stripes; those that don't, don't, regardless of lineage.

Obviously this doesn't prove anything about spiders. But it does tend to undermine the assumption that spotty distribution implies mere random variation. In these two cases, it indicates a broadly distributed latent capability that can be quickly switched on or off by natural selection as needed.

Assuming PZ's hypothesis is true, and the variation is not due to selection, what would be the proper term for that? Nonaptation? Spandrel?

Technically, a spandrel is not just a selectively neutral trait. It's a selectively neutral side-effect of some adaptive trait. For example, the male nipple is (arguably) a spandrel, since it's a non-functional side-effect of the process that produces female nipples.

So to argue that spider fluorescence is a spandrel, you'd have to identify an adaptive trait that produces fluorescence as an incidental side-effect.

By Gregory Kusnick (not verified) on 12 Dec 2007 #permalink

Great post PZ. I have to disagree with your hypothesis because looking at the taxonomic distribution of florescence, we can (obviously) see that those species that: spend time in flower heads, build webs high in foliage, build webs across exposed surfaces, all have high values while those groups whose species are ground dwellers, make webs concealed in vegetation, or scour plants- patrolling for prey- all of these groups do not exhibit the brightest level of florescence. This is a bit of an oversimplification. However, consider the benefits of attracting and capturing pollinators which are themselves attracted to the UV reflectance of flowers.

On a related topic, about half of the mite species (Acari) are florescent. Here is a place where the evolutionary reason is not apparent- but that just means we have a lot to learn, not that the property is necessarily incidental.

Re #25

Thanks Gregory!

Is there a term for things that are selectively neutral but are not side-effects of adaptive traits? Anyone know any examples of such things?

Would it be correct to say, then, that nonaptation is a more general term, and that spandrels are just one category of nonaptation?

These are 'true' spiders that were tested here. I'll have to go home and try this out on tarantulas and see if they exhibit the same thing. Might make it easier to find em when they go hiding in their webs and what not. Thanks for the post.

Actually, I've heard some things about experiments with where spiders use the fact they are bright in UV (can't find the paper, but I know what spider it is) This white spider: http://www.dkimages.com/discover/previews/929/45001451.JPG - eats bees. And it catches them by sitting on flowers and waiting for the bees to come to the flower.

Bees can see in UV. While the spider is white to us, it's not white to the bees and it actually glows brightly in UV, brighter than even white flowers (though it does rest on other colored flowers).

Turns out, that the bees actually prefer flowers on which the spider is resting. So, camouflage? Not exactly. More like lure, least in this case.

Off-topic, but since someone brought it up:

How strong is the latitude-skin color correlation really? Eskimos, Mongolians, I'm looking at you. Are hereditary Incas darker-skinned than patagonians?

By YetAnotherKevin (not verified) on 12 Dec 2007 #permalink

I have an LED headlamp and I was surprised to see hundreds of glowing dots in my backyard. Upon closer inspection most of them were spiders. It was quite eye opening to realize the sheer density of spiders in my backyard. Something you don't usually notice.

By Casey Schmidt (not verified) on 12 Dec 2007 #permalink

Very cool. Yeah, I can't help wondering why certain families seem to display more fluorescence, but perhaps it's a random byproduct of something else.

The two families up there with the best visual acuity are probably Salticidae and Lycosidae, but it seems that spiders in Araneidae, who have comparatively poor vision, are winning out in the fluorescence department, which to me suggests that perhaps it doesn't serve a communicative or courtship function within the species. Of course only a few species from each family were sampled; it would be great to have more data.

Of course, in terms of testing spiders' visual acuity, I'm also not sure how much attention has been paid to their perception of wavelengths not visible to humans.

I volunteer to play around with black lights and spiders.

How strong is the latitude-skin color correlation really? Eskimos, Mongolians, I'm looking at you.

It is pretty strong, though of course there's also phylogenetic inertia (in other words, evolution takes time), and other effects -- eskimos traditionally get vitamin D from their diet, so they don't need extra-transparent skin, for example.

Are hereditary Incas darker-skinned than patagonians?

AFAIK yes, though people from the Amazon rainforest are darker still.

By David Marjanović, OM (not verified) on 12 Dec 2007 #permalink

YetAnotherKevin, this blog has several fascinating posts on the genetics and evolution of skin color. The link I gave will bring you to one of them, and the search field on the left will lead you to many more.

By albinosquid (not verified) on 12 Dec 2007 #permalink

spiders and scorpions are both arachnids. not terribly surprising that they share this property.

By arachnophilia (not verified) on 12 Dec 2007 #permalink

Every fall, we get some really nice orb-weavers setting up shop on or around our porch. I noticed the fluorescence this past *spider season*, when I got out my trusty blue LED flashlight one evening. We get rather attached these critters, and they have become part of our seasonal observances.

By Kerry Maxwell (not verified) on 12 Dec 2007 #permalink

I recall that some orb web spiders weave thicker patterns in their webs that are highly visible to bees. I remember because I always assumed that these zigzagging thicker parts of the web were repairs of damage made to the web by previous insect catches or possibly windblown debris. It was very cool finding out that this behavior was actually a part of the spider's feeding strategy. Have you ever seen one of these spiders sitting smack in the middle of an "X" pattern of these zigzaggeries? Maybe there is a correlation between the amount of flourescence in spider chitin and this particular predation strategy?

By TritoneSub (not verified) on 13 Dec 2007 #permalink

Having tried a whole lot of fluorescent and LED light sources, one recommendation -- my favorite UV flashlight; single AAA alkaline:

http://www.arcflashlight.com/arc-aaa-uv.shtml

Tiny (handy on a keychain/reel/belt clip)
Long battery life. Very sturdy.

By Hank Roberts (not verified) on 13 Dec 2007 #permalink

I suspect PZ is right that that the glowy stuff is in spiders more or less by accident, but can then be co-opted in an adaptive way. Without clicking through to the piccie Kimberly @29 links to, I am going to bet it is M. vatia or at least a related Thomisid. And I've also read of studies showing that UV fluorescence actively lures prey.

But the Dysderidae are also brightly fluorescent. These six-eyed spiders -- they have whacking great fangs that they use to pierce woodlouse armour -- don't see well, and in any event live underneath things. Hard to see how glowing under UV illumination would be of any help to them. It's intriguing that they are, as Prillatashekta notes @8, pretty primitive and hence perhaps less likely to have found a use for their cool UV trick.

In one of the above comments it seems like someone is trying to distinguish neutral side effects of adaptive features from the neutral side effects of non-adaptive ones, calling the former spandrels. But how can something be the side effect of a neutral feature? How does one tell the direction of the hierarchy? Seems like if both are phenotypic then there is no way to tell which is the effect and which the side effect. Is having a boundary a side effect of occupying finite space or is occupying finite space a side effect of having a boundary.

Seems more natural to think of replicators as being the effects and the phenotypic results as being the side effects. But this kind of throws a wrench in Gould's works since he wants to treat phenotypes as being side effects of other phenotypes.

If spiders blood just happens to glow because the best chemical hit upon by a random walk happens to glow that doesn't necessarily mean that a better chemical doesn't exist that fails to glow. So is it truly a spandrel?

Since spandrels are side effects of adaptive phenotypic features, and genes themselves are both genotypic and phenotypic adaptive features (as is the ability to mutate) then aren't all non-adaptive features spandrels at some level or another.

Since some are allowing for inter-individual definition of spandrel (nipples on men and perhaps vaginas in lesbians. Aren't many neutral mutations a natural side effect of the ability of our genetic material to have sex, an adaptive feature? ;) I sure wish Gould was clearer in his definitions. A little more scientific rigor and a little more thought might have cleared up such speculation.

So does anyone have a clear idea of what a spandrel is? Must the side effect of the feature exhibit itself in the same organism? Cute little ducks seem to have the side effect of inspiring cooing in humans. Is my emotional reaction to cute animals a spandrel or my appreciation of flowers?

If a feature originally adaptive flips in relationship with it's side effect, and becomes the spandrel then is it truly a spandrel. For instance if domes no longer serve the purpose of shelter because perhaps we grow fur, but instead are kept around, or built, for the beautiful displays on their spandrels then does the dome become the side effect of the spandrel.

If we consider neutral side effects to be spandrels then why not also non-adaptive side effects. There is nothing to prevent a side effect that is negatively selected from becoming an asset with a change in environment or genetics.

Do probabilistic features count as spandrels? Some adaptive features might have the side effect of increasing the odds of some other feature occurring as mitigated by environmental or chance circumstances. If for instance cancer were actually a chance side effect of some adaptive feature of our cells then one might then ask is cancer a spandrel. Well you say - no because it is deleterious. However, with environmental change it is not. Think of the poor Tasmanian devil, or of eternal cancer cell lines adapted to live in laboratories.

In this speculative scenario what was once a deleterious side effect now becomes an adaptive feature. If this doesn't count as a spandrel it certainly seems to me to be as important a concept as spandrels. The bad effect need not be lethal just so long is it is out weighted by the adaptive effect of the base feature. Therefore such effects may be quite common.

Now all I have to do is coin a name for my new and very important discovery of biology. ;) Perhaps I will call it a Rube-Gouldberg.

Heck what about an anti-adaptive feature that has the side effect of leading to an adaptive one. Pimples on teenagers making them breed later in life yet causing them to have an overall higher rate of reproduction because in certain environments they don't produce more kids than they can feed.

Another example, nerdy anti-social traits leading one to spend many hours on geeky pursuits like studying that leads to success in later life.

By Brian Macker (not verified) on 13 Dec 2007 #permalink

Hm. That spider-fluorescence might be a spandrel is interesting. If humans started using UV as a cleaning aid, it'd rapidly become maladaptive, at least for populations of spiders living inside dwellings.

What would happen if you left the blacklight on the spider for an extended period of time? Specifically a tarantula?

How strong is the latitude-skin color correlation really? Eskimos, Mongolians, I'm looking at you.

It is pretty strong, though of course there's also phylogenetic inertia (in other words, evolution takes time), and other effects -- eskimos traditionally get vitamin D from their diet, so they don't need extra-transparent skin, for example.

Are hereditary Incas darker-skinned than patagonians?

AFAIK yes, though people from the Amazon rainforest are darker still.

By David Marjanović, OM (not verified) on 12 Dec 2007 #permalink

I need to tell you spiders are weird and want yall to know that I just found a green glowing spider.

By Kelsea May (not verified) on 28 Mar 2009 #permalink