Myths in Biology: Mendel's Pea Flowers

At P.Z. Myers' blog, Pharyngula, he has a post on the subject of "myths" in biology, using the number of cell types in humans as an example. He cites Stephen J. Gould's famous article on fox terriers as an example of such myths and how they get incorporated in the science of biology.

You want myths in biology? Pick up any introductory biology textbook and look up Mendel’s original experiments with garden pea plants. Look at the color illustration of the seven phenotypic characters Mendel supposedly tested. Is purple versus white flower color in the figure? Now, read Mendel’s paper describing the seven different characters in garden peas that he studied (you can find the original paper here, in the original German and in English translation). Is purple versus white flower color in the list of characters tested? Interesting…

Toward the end of the paper, Mendel mentions that in a later set of experiments he tested flower color and found the same ratios that he found with the original seven traits he tested. According to the paper, he tested “violett-rothe und weiss Blüthenfarbe” (i.e. “violet-red and white blossom color”), but this test was NOT in his original set of seven experiments, which are the ones always illustrated in biology textbooks.

So, when did biology textbooks start this particular myth? As far as I can tell it was in the first biology textbook with full color illustrations: William T. Keeton’s Biological Science, 2nd ed. The illustrator thought he could kill two birds with one stone by illustrating the technique used to ensure controlled fertilization (i.e. removing the stamens from the flowers using iris scissors) and purple and white flower color. Except that Mendel didn’t study purple versus white flower color in his original series of seven crosses.

Which pair of traits did Mendel actually study, but were replaced by purple versus white flowers in all introductory biology textbooks? The color of the seed-coat, in which gray-brown is dominant and white is recessive. According to Mendel’s original paper, gray-brown seed coats are associated with (what we would now refer to as linked with) “violet-red blossoms and reddish spots in the leaf axils,” but once again Mendel did NOT explicitly test purple versus white flower color in the experiments for which he is remembered, and for which the science of “Mendelian genetics” is named.

And why is the clearly incorrect list (and colored figure) of the seven traits Mendel supposedly studied included in every introductory biology textbook today? Because Keeton’s textbook was the most widely used textbook in biology for decades, so all of the other publishers simply copied what was in his textbook as a way of gaining market share.

Can this myth be corrected now? How many professors’ sets of lecture notes and PowerPoint slides would have to be changed to correct this mistake, and how many textbooks would need new illustrations that included the correct list of the seven traits, and how many people would complain about these changes, or (even worse) suggest that Mendel really did study purple and white flower color in his original series of seven experiments?

************************************************

As always, comments, criticisms, and suggestions are warmly welcomed!

--Allen

Why vaccinate your children?

QUESTION: Not immunizing does not expose someone to a disease, only actual exposure to the disease does so. If a vaccine is effective and your children are vaccinated, how does an unvaccinated person put them at risk? Especially if that person has not been exposed to the disease in question?

ANSWER: It has to do with herd immunity (also known as "community immunity"). With the exception of the smallpox vaccination (reaction to which is easily observed), none of the vaccinations we give our children are tested individually to see if they "took". This is difficult at best and next to impossible in many cases. Instead, the effectiveness of a vaccination program is measured by the prevalence of the disease before, during, and after vaccination of a target population. If the prevalence of the disease goes down as more people are vaccinated (this is verified using statistical analysis), then the vaccination is assumed to work even if it doesn't in some cases.

In fact, no vaccination (including the smallpox vaccination) works in every single case. There is a small, but non-zero fraction of any vaccinated population in which the vaccination doesn't "take" (this again is statistically testable). What this means is that in any population that has been vaccinated, there is a small residual fraction that is still susceptible to the infection.

The rate at which an infection is transmitted depends upon the "infection triangle": (1) virulence (how easily the infectious agent enters a potential host), (2) resistance (how easily the host fights off the infection), and (3) prevalence (how many carriers/potential spreaders there are in the population). Herd immunity depends on 2 and 3, and is defined as that percentage of a vaccinated population that is high enough to stop further transmission of the infection. For example, epidemiologists (doctors who study how infectious diseases spread through populations) define the "herd immunity threshold" for diptheria as 85% (see here). This means that in a population in which at least 85% of the individuals have been vaccinated against diptheria, the probability of an infectious individual spreading the disease to an unprotected member of the population (i.e. the 15% who have not been vaccinated or in whom the vaccination hasn't "taken") is sufficiently low that unprotected individuals probably won't be exposed.

The problem here is that "unprotected" doesn't necessarily mean "unvaccinated". Like any real process in the real world, not all vaccinations "take". A small but non-zero fraction of vaccinations don't cause the vaccinated individual to develop permanent resistance to the disease. This can happen from a number of factors, including the immune status of the vaccinated individual and variations in the potency of the vaccine. This means that, like the actual rate of resistance following vaccination, the probability that a vaccinated individual is actually resistant to the infection is not 100%, but usually somewhere between 80% and 95%. What this means is that the actual level of herd immunity is lower than the rate of vaccination, sometimes by quite a bit (this is a function of both individual variations and group variability - some people are more likely to develop resistance than others).

Therefore, if people in a population don't get vaccinated, then the size of the susceptible carrier population is larger, and if it's large enough (i.e. greater than the "herd immunity threshold"), then the probability that the infectious agent will be spread is high enough that susceptible individuals (i.e. those that have not been vaccinated AND those whose vaccinations didn't "take") will be exposed to the infectious agent, get the disease, and spread it to other susceptible individuals.

This is why there have recently been mini-epidemics of pertussis (whooping cough), measles, and other infectious diseases that were formally almost completely eliminated. Badly educated people (including, but not limited to people who have a visceral anti-science/anti-government bias) don't vaccinated their children, who become part of the susceptible population who, if they are numerous enough, can spread the disease to others, including those whose vaccinations didn't "take".

So, how do you know if your vaccination (or your kid's vaccination) "took" or not? Simple answer: you don't (indeed, in most cases, even with the old smallpox vaccination, you can't). Therefore those of us who don't want our children sickened, crippled, or killed by a preventable infectious disease are counting on everyone else getting their children vaccinated to the point at which the "herd immunity" of our community is above the "transmissibility threshold" and therefore won't be exposed to the infectious agent.

The same kind of reasoning that underlies the concept of "herd immunity" can be used to see if the argument that increased vaccinations cause autism is valid. If there is a causative relationship between increased vaccinations and autism, and if the underlying causative agent is thimerosol in the vaccinations, then there should be a decrease in autism since thimerosol was removed from the vaccinations. No such correlation has been found; ergo, thimerosol in vaccinations was not the causative agent for the perceived increase in frequency of autism.

Another potential test of this hypothesis is to see if those children who, for whatever reason (e.g. immune dysfunctions, religious prohibitions, lack of access to vaccines, etc.) have NOT been vaccinated have lower rates of autism. This is also not the case, so once again the hypothesis that vaccinations (and specifically vaccinations with serum that has been preserved with thimerosol) cause autism is unfounded. This is precisely the kind of statistical correlation testing that is the basis for every single one of the peer-reviewed epidemiological studies that have been done to determine if there is a causative link between vaccinations and autism (or ADHD, or your choice of mental disorder for which we do not yet know the underlying cause). And every single one of these studies (and there have now been many, including several meta-analyses of multiple studies) have shown no statistically significant (i.e. real) correlation between rates of vaccination and the observed rate of increase in autism, ADHD, and some other developmental disorders?.

So, what is causing the statistically detectable increase in the rate of autism, ADHD, and some other developmental disorders? There are plenty of candidates. People (both men and women) are having children at older ages, which has been shown to be causally related to some of these disorders. People are exposed to increasing levels of artificial chemicals and heavy metals in the environment, which again have been shown to be causally related to some of these disorders. Perhaps most significantly, the diagnostic criteria for some of these disorders have changed, especially for ADHD and autism, which are now considered to be "spectrum" disorders, rather than single pathologies that one either has or does not have. Had the current criteria for these developmental disorders been used when I was a kid, I would almost certainly have been diagnosed with ADHD and also probably mild Aspergers' syndrome (i.e. autism spectrum disorder). When you change the definition of a disease, you change your perception of its prevalence.

The same is not the case for the demonstrated decline in crippling and potentially fatal diseases, however. There is very strong statistical evidence that this dramatic decline over the past century and a half has been due to two factors: vaccination and public health. Both of these were and are developed by medical scientists, implemented by health professionals, and supported (and in some cases legally mandated) by governments, usually at the state level.

I believe that it is the case that in New York State you can refuse to have your children vaccinated for religious reasons, yet the public schools are still required to let those children attend classes and therefore expose all of their classmates to an increased possibility of contracting a potentially crippling or fatal disease. In my opinion this is wrong: children who have not been vaccinated should not be allowed to even enter a public building, much less attend public school. Sure, their parents can exercise their right to endanger their children's health (whether their children have a right to not be endangered by their parents is another question), but that right should not take precedence over the rights of other parents (and their children) to not be exposed to the threat of contracting the diseases that such parents have exposed their children to.

************************************************

As always, comments, criticisms, and suggestions are warmly welcomed!

--Allen

The Evolution of Cooperation – Three Theories or One (or None)?

Ever since Darwin, evolutionary biologists have thought and written (and argued about) the evolution of cooperation. In the 20th century, at least three different major theories were proposed to explain how cooperation (and especially unselfish altruism) could evolve by natural selection. Several attempts to unify these theories have been made and are currently a topic of intense debate.

Now, as part of Cornell's Darwin Days celebration, you are invited to come to a dinner discussion to hear about, think about, talk about, and (hopefully) argue about these theories and their implications for human behavior, ethics, and philosophy.

PRESENTER: Allen MacNeill, Senior Lecturer in Biology and Evolution at Cornell and author of Evolutionary Biology and Evolutionary Psychology

SPONSORS: Liebermania! and the Cornell Human Ethology Forum (this will also be the organizational meeting for CHEF)

DAY, TIME, & LOCATION: Thursday 14 February 2013 at 6:00 PM at Risley Dining/Tammany at Cornell University. This event is part of Cornell and PRI/MOTE's Darwin Day 2013 celebration.

Please contact Allen MacNeill at adm6@cornell.edu by Tuesday 12 February 2013 if you plan to attend this event. Hope to see you there!