Spider Tales 3: The Widow and the Rattlesnake... And other LD50 stories

 "A black widow's venom is 15 times more toxic than a rattlesnake's" 
                                                                            National Geographic documentary

It's definitely one of the most commonly shared pieces of spider trivia. It's hard to escape it: every time a wildlife documentary features a black widow, you can bet this comparison with rattlesnake venom will pop up at some point.
You may be surprised, then, to learn that this claim doesn't say much about how "deadly" those spiders are. Actually, it doesn't say much about anything: although it does sound a lot like a fact, it's just a meaningless factoid, little more than a mere urban legend.

Venom 15 times as potent as a rattlesnake's? That sure sounds scary... But what does it mean?

Similar comparisons, based on LD50 values, are also often shared about some species of venomous snakes (particularly some Australian Elapidae famous for being "the most venomous snakes on Earth") and some other venomous or poisonous animals. They do sound scary, and make these animals sound very deadly. But are they actually informative?

"Drop for drop, boomslang venom is the most potent of all African snakes" is another common factoid based on LD50 comparisons

What is LD50?

Venoms and other toxic substances can have an enormous variety of different effects on living organisms. Some overstimulate or shut down nerve receptors, causing a wide array of neurological issues such as extreme pain, involuntary muscle contractions, heartbeat irregularities, or paralysis. Some thin or thicken the blood, causing anaemia, haemorrhages or clotting. Some cause the destruction of cell membranes, which results in tissue necrosis in skin or internal organs, or hemolytic anaemia. Some toxic substances even alter the DNA or mess with the way we assimilate some vital chemical elements. 

The effects of toxic substances vary from barely noticeable to lethal, and how fast their effects appear also varies a lot. Some have perceptible effects almost immediately upon exposure, while others can take months or even years of prolonged or repeated exposure before the first symptoms of poisoning appear.
It all depends on the type of poison, and, even more importantly, on the dosage. With toxic substances, it's all about the dose. Anything, including elements necessary to our survival, is toxic in large enough amounts: even water poisoning is a thing!
However, some things are more toxic than others, and some can cause death or harmful effects even in minute amounts.
How much is needed to cause observable poisoning, or death, is what measurements such as lethality and toxicity tests strive to show.

In the broad lines, these tests are simple: different, and increasingly large, doses of the tested substance are administered to subgroups of a sample population of animals or cell cultures, until the investigated effect is observed in all the individuals of a subgroup; that's how the doses of substance required to cause these effects are determined¹.
In lethality tests, the effect in question is the death of the individuals. The LD50 is the minimal dose, in mg of substance per kg of organism, that kills 50% of the sample population¹ (LD50 means "Lethal Dose 50%"). The more potent the substance, the lower the LD50 value will be.

The venoms of some cone snail species (thankfully not this Conus ventricosus) have some of the lowest LD50 ever recorded in venomous animals (NEVER TOUCH OR HANDLE A LIVE CONE SNAIL)*

Of course, it is extremely useful to know how lethal a substance can be. It can, for instance, show how effective some potential pesticide could be against a targeted "pest" species, and how harmful it could be for those we don't want to harm.
It's also really useful in the early stages of the pharmacological study of a molecule, in order to assess whether or not its toxicity is likely to be a problem for use as a drug.
In venom studies, these LD50 measurements can provide a rough idea of how potent the venom of a species is to mammals, including humans. 

Generally, lethality tests (when it's the toxicity to humans that is investigated) are conducted on 20 grams mice.

A measurement to be taken with MANY grains of salt

It is, however, very important to always keep in mind that LD50 values are obtained from tests on mice or other non-human animals.
Toxicity is relative, and the same substance often affects different species differently.
It is, for instance, a well-known fact that some mushrooms are so toxic to humans that a few grams are enough to kill an adult. Meanwhile, slugs eat those mushrooms, apparently unaffected by their toxins.
Mediterranean black widow (Latrodectus tredecimguttatus) venom has been shown to be considerably more toxic to mice than to frogs, much more to flies than to cockroaches, and that guinea pigs are much more sensitive to it than mice. Injected to guinea pigs, that venom is more than 1000 times as potent as... Itself, injected to a frog! 
Even genetically different stocks of mice can have different sensitivities to the same substance¹, so precise LD50 comparisons aren't really relevant if the animal species (or even, ideally, the genetic stock) on which the tests were conducted is not specified. 

That's also why LD50 can only give a rough idea of the potential toxicity of a substance to humans, and should never be extrapolated directly from mouse to human.
Brown button (brown widow) spider venom (Latrodectus geometricus), for instance, was found to be much more lethal to mice than Northern (Latrodectus variolus) and Southern black widow (L. mactans) venoms, in the same experimental conditions. There is, however, no indication that it is also the case for humans; actually, envenomations on humans by L. geometricus are, on average, markedly less severe than bites from black widow species (but, as we will see later, there are other factors involved).

The opposite is also true: for instance, Sydney funnel-web spider (Atrax robustus) venom is much more toxic to primates, including humans, than it is to other mammals such as rats, mice or cats.
Therefore, throwing LD50 values around when talking about venomous animals is entertaining, but not very useful. Saying that something is "highly toxic" without specifying "highly toxic to what" is not informative, and comparing different measurements of lethality to mice like they're directly applicable to humans is actually misleading.
Again, as mice and humans are quite closely related, something that's very toxic to them is likely to also be harmful to humans, but saying some species of snake or spider is "twice (or fifteen times) as venomous" as another, because the LD50 of its venom is twice lower than the other one, actually makes little sense.

Brown button spider (Latrodectus geometricus) venom was found, in a comparative study, to be much more lethal to mice than venom from other widow species, but this doesn't mean much about its toxicity to humans

Another thing about lethality tests is that they're not only highly influenced by the species they're tested on, but also by the way the substance is administered. Different methods of delivery can be used in the tests.
In lethality experiments, venom is most often injected intravenously (directly in blood vessels), in muscles (intramuscular), under the skin (subcutaneous), or through the peritoneum (intra-peritoneal).
Depending on the type of venom and its main effects, the differences in LD50 induced by these different methods of injections can be vast: administered through intra-peritoneal injection, western diamondback rattlesnake (Crotalus atrox) venom has a LD50 26 times lower (0.72 mg/kg) than... The same venom, injected intramuscularly (19.04 mg/kg)!
That's one of the most important reasons why the "black widow and rattlesnake" comparison makes no sense: while it can be safely assumed both of the compared LD50 were obtained on mice, there's no way to guess how the experiments that yielded these values were conducted.
If the experimental conditions are not specified, LD50 comparisons are completely meaningless. 

In addition to that, the "15 times more potent than rattlesnake venom" factoid makes things even worse by not mentioning which exact species of "black widow" and "rattlesnake" are compared. Both black widows (genus Latrodectus) and rattlesnakes (genus Crotalus) are not species, but genera, i.e. groups of related species.
There are 34 different species of widow spiders, most of them black and called black widows (African species are locally known as "black button spiders"), and about 50 currently recognised rattlesnake species; and different species have different venoms.
In the same experimental setting, significant differences in lethality between venoms of different Latrodectus species have been shown.
Rattlesnake venoms vary even more in terms of effects and lethality from one species to another, sometimes even one subspecies to another. Injected intravenously to mice, the LD50 of the South American rattlesnake's venom, Crotalus durissus, is 113 times lower than the Tancitaran dusky rattlesnake's (Crotalus pusillus).

"Venom 15 times more toxic than a rattlesnake's"? Which "black widow" species? Which rattlesnake species? Toxic to what? In what experimental conditions? Without these crucial elements, it doesn't mean anything...

Without specifying the exact species and the mode of injection, it would be just as easy to cherry-pick LD50 values showing black widow venom "is 15 times more potent than a rattlesnake's" (for instance, by comparing the LD50 of Mediterranean black widow venom injected intraperitoneally, 0.59 mg/kg, with Crotalus atrox venom injected subcutaneously, 7.8 mg/kg) as it would be to "prove" the opposite (by comparing the same black widow LD50 with C. scutulatus venom injected intravenously, 0.03 mg/kg)!

Interestingly, when LD50 values with the same modes of injection are compared, our factoid seems to be mostly false: in the experiments conducted by McCrone in 1964, intraperitoneal LD50 on mice was 0.59 mg/kg for the Mediterranean black widow (Latrodectus tredecimguttatus), 0.43 mg/kg for the brown widow (L. geometricus), 1.30 mg/kg for the Southern black widow (L. mactans), 1.80 mg/kg for the Northern black widow (L. variolus) and 2.20 mg/kg for the red widow (L. bishopi).
These values are in the same range as the intraperitoneal LD50 of many rattlesnake venoms, such as the eastern diamondback (Crotalus adamanteus, 1.67 mg/kg), the western diamondback (C. atrox, 0.72 mg/kg), or the northern Pacific rattlesnake (C. oreganus, 2.3 mg/kg), and much higher than the South American rattlesnake (C. durissus, 0.0478 mg/kg), whose venom seems to be one of the most potent (to mammals) among rattlesnakes.
That claim is only somewhat true with the highest intraperitoneal LD50 value among the sampled rattlesnake species, Crotalus willardi (12 mg/kg).

Contrary to popular belief, black widow venom is not much more potent than the venom of most rattlesnake species

These issues also apply to similar claims about other venomous animals, such as "the inland taipan (Oxyuranus microlepidotus) has the most toxic venom of all snakes", "the Brazilian wandering spider (Phoneutria nigriventer) is 16 times more venomous than the black widow" or "drop for drop, the boomslang has the most potent venom among African snakes".
These claims generalise something which can not, by definition, be generalised. LD50 values only make sense within the experimental setting in which they are obtained. Depending on how the venom is injected, and (of course) to what, each of them can be true or false.
They are, therefore, worthless, and do not achieve anything besides sounding scary. 

However, even if the right information was added to these assertions to make them correct, they would still not say much about how dangerous to humans these animals actually are. It takes more than just potent venom to make a dangerous species.

Venom potency and danger

Steel is about 8 times denser than the average watermelon, and much harder.
It would be quite silly, though, to understand that fact as "steel hurts more than watermelon". Of course, a 5 kg watermelon, hurled at a human face, will hurt much more than a 5 g steel ball launched at the same speed. Sure, steel is harder and denser, but that doesn't really matter when the size difference is so great.
Same goes with venom: potency is one thing, but one thing that matters much more is the dose. 

Even if the comparison between black widow and rattlesnake venom actually made sense, it would absolutely not mean that a black widow's bite on a human would be more dangerous than a rattlesnake's. At all.
Why? Because the volume of venom a rattlesnake injects when it bites is several orders of magnitude greater than what a black widow delivers.
In one defensive bite, a Western diamondback rattlesnake (Crotalus atrox) typically delivers 3 to 28 mg of venom per fang, i.e. 6 to 56 mg if both fangs penetrate the skin. Meanwhile, an adult female Western black widow (Latrodectus hesperus) will typically inject 5 to 10 µg (0.005-0.01 mg) of venom in a defensive bite.
Therefore, a rattlesnake typically injects a volume of venom roughly 1000-10 000 times larger than what a black widow does.
Thus, even if it was indeed 15 times less potent than a black widow's, the sheer amount of venom injected by a rattlesnake (at least by one of the larger species of rattlesnakes) would still make its bite much more toxic than a black widow's. 

In addition to that, widow spiders have short fangs, are web-bound and only bite humans if pressed against the skin. Statistically, a close encounter between a human and a black widow is therefore very unlikely to result in a bite. The people unlucky enough to get bitten and envenomated are generally in for an excruciatingly painful, exhausting and scary experience that will last a few hours to several days (and sometimes lingering after-effects for weeks), but even without antivenom, the prospects of survival and eventual complete recovery are very high (96-99.8%). Antivenom or not, fatalities are extremely rare.

Rattlesnake bites, on the other hand, are not only more likely to happen in case of close confrontation between human and snake (although, like any snake, it will take any opportunity to escape without a fight if possible), but are also a much more serious medical emergency.
While the effects of black widow venom are almost purely neurotoxic in nature (it affects the nervous system) and reversible, rattlesnake venom (the intensity of the different effects vary depending on the species) is neurotoxic, haemotoxic (messes with blood coagulation) and cytotoxic (it destroy tissues). Envenomations by the larger species almost always require multiple vials of antivenom to counteract the effects of the venom, and can cause permanent damage to skin, limbs and internal organs (most notably kidneys), particularly if the antivenom is administered too late. 

Therefore, less potent venom or not, rattlesnakes are much more dangerous animals than black widows; there is no debate about that.

In case of close encounter with humans, widow spiders will try to avoid wasting their venom in a defensive bite; those generally happen when a spider finds itself accidentally trapped against the skin*

Lethality and dangerousness are actually two independent notions.
Lethality is the ability of a substance to cause death, but death is only one of the negative consequences (arguably, the most extreme) an envenomation can have.
Just because the venom of an animal species is rarely or never deadly to humans (or not as lethal as the venom of another species), doesn't mean it can't cause some harrowing, or even incapacitating symptoms.
A non-fatal bite from a venomous snake can still be an ordeal, sometimes causing irreversible damage to skin, limbs, internal organs or parts of the nervous system; it can even result in permanent invalidity. Even when the effects are reversible, victims can find themselves temporarily unable to work and provide for themselves and their families, and, in some cases, have expensive hospital costs to cover; it can therefore be economically devastating to people who were already in precarious financial situations.
In countries with very high prevalence of snakebites, their impact on public health and economy is significant, even though the vast majority of victims survive the bites.

Venom lethality of the involved snake species (as long as the venom in question is potent enough and injected in large enough quantities to cause severe envenomations in humans) actually does not matter much in the picture of snakebite significance.
While Australia is famous for being home to many of the "most venomous snake species on the planet" (in terms of LD50), it is also one of the places on Earth where the prevalence of both snakebite cases and fatalities are the lowest. 

On the other hand, many of the world's most dangerous snakes, in terms of actual impact on public health (in number of bites on humans, likelihood of an encounter with humans, and of this encounter resulting in a bite, and common consequences of an envenomation) are far from being among the most potent in terms of LD50 on mice.

For instance, in South and Central America, the snakes armed with the most toxic (to humans as well as mice) venom are, arguably, coral snakes (genus Micrurus); human envenomations by coral snakes are very dangerous and can be life-threatening, the neurotoxic venom causing neuromuscular blockade, resulting, in severe cases, in paralysis and respiratory failure.
However, the most dangerous and feared snakes on the continent are not coral snakes, but pitvipers, particularly species in the genus Bothrops, known as "lanceheads" or "fer-de-lance".
While their venom is less potent (particularly in terms of subcutaneous LD50) than most coral snakes, Bothrops spp. have long fangs, large venom glands, cryptic colours, and a tendency to freeze and rely on camouflage instead of fleeing from potential threats, and strike readily. Rodents make up a large part of their diet, and because of that, they can be abundant in agricultural zones and around human settlements.
Although fatalities are rare, and only a minority (4-14% of envenomation cases, depending on the species) of bite cases are severe, Bothrops bites can still result in coagulopathy, haemorrhaging, necrosis, acute kidney injury, and are prone to infection.
That combination of traits makes Bothrops species the most dangerous snakes in South America, as they are the most commonly involved in snakebite cases: in Brazil, up to 70% of bites by venomous snakes are caused by this genus!
Meanwhile, coral snakes, with their bright colours, shy nature and tendency to frequently administer dry (venomless) bites, only account for 1-2% of snakebite incidents; despite their more potent venom, they are, therefore, less dangerous than lanceheads.

While their venoms aren't the most potent among South American venomous snakes, their abundance, camouflage, large fangs, and tendency to stay still and strike readily when threatened, make Bothrops species (like this B. atrox) the most dangerous snakes on the continent (photo: Jérémie Lapèze)

A similar situation exists in Southern Africa: the boomslang (Dispholidus typus) has a venom with a lower (intravenous, mice) LD50 than the puff adder (Bitis arietans) or the Mozambique spitting cobra (Naja mossambica), and even (again, only intravenously) than the black mamba (Dendroaspis polylepis). However, bites by this shy, fast, mainly arboreal species are quite rare, and are most often induced by attempts at handling the snake.

Meanwhile, although not the snakes with the most lethal venom in the region, the puff adder and Mozambique spitting cobra are the most common cause of severe bite incidents, despite sharing the area with 15 other dangerously venomous snake species!
In addition to their very toxic venom that can cause extensive tissue damage, these snakes display behaviours which make them susceptible to conflict with humans: the Mozambique spitting cobra has a nervous, highly defensive disposition towards potential threats²; the puff adder often freezes in its tracks and relies on camouflage when approached² instead of fleeing, and can strike with impressive speed². In addition to that, these species are fairly often found at ground level² in the vicinity of human dwellings and farms, attracted by rodents, and the Mozambique cobra is even known to sometimes enter houses at night.
The black mamba (Dendroaspis polylepis) causes much fewer bites, but its large size, nervous temperament and its ability to inject large doses of highly neurotoxic, fast-acting venom which makes its bites frequently life-threatening, make it a very dangerous animal. It is often regarded as the most dangerous snake on the continent², and is indisputably much more dangerous than the boomslang.

Its venom may not be as potent as other species in the area such as the black mamba or the boomslang, but a puff adder's bite can still cause extensive, sometimes permanent, and potentially life-threatening damage to the human body, and its camouflage and behaviour make it susceptible to conflict with humans

Therefore, animals with extremely lethal venoms are not always the most dangerous: many factors matter more in the picture than just venom potency!
Much more than the drop-for-drop toxicity, the ability of the animal to efficiently deliver large quantities of venom, the propensity to do so in case of confrontation, and the likelihood of encounter and hostile interaction with humans, are what makes the dangerousness of a venomous species.

Lethality tests can be very useful for scientific research, for multiple applications, and in that context, LD50 values definitely matter.
From a nature enthusiast's perspective, they can be entertaining and fun to talk about and compare, and the fact that some animals have evolved such impressively efficient weapons definitely inspires a sense of awe and respect.
They are not, however, a tool for understanding, and even less for precisely measuring, how dangerous to humans the animal that carries the venom would be in the field. This is not their purpose, and they should never be understood as such.
In addition to that, they make very little sense when taken out of their context, and comparisons of values obtained from unspecified experimental settings will only be empty, meaningless factoids; all these can do is mislead and spread unnecessary fear.⏹

* I am a trained professional, experienced in working with medically significant arachnids and other venomous wild animals in the field. Please do not attempt to replicate what you see on these images; free handling wild animals, particularly those potentially capable of harming humans, is never advised, and is not something I do or condone outside of specific circumstances.

References

References are integrated in the text of the article; the words in blue are clickable and will redirect you to the sources of the information.

Paper references:

¹ Rollard C., Chippaux J-P. & Goyffon M., La fonction venimeuse. Ed. Lavoisier, Paris, 2015.
² Alexander G. & Marais J. A Guide to the Reptiles of Southern Africa. Ed. Struik Nature, Cape Town, 2007.

Except when the source is explicitly cited, the images illustrating this blog are mine and are not free to use.
The images by Jérémie Lapèze are used with the author's consent and are not free to use.