Question
QUESTION: Hi Melissa,
I have a sampler from New England, done around 1820, and it has a lot of plants depicted on it. I have a theory that it is actually a kind of catalog of medicinal herbs. I've had experts look at the sampler and they've identified about 15 species. All of them check out as having medicinal properties. Only one of them is poisonous. My question is, could this be a random result? If you randomly picked a list of 15 plants, would it be typical that only one of them is poisonous, and all of them had medicinal uses in 1820? Or is that unusual? It's my impression that a much higher percentage of plants are poisonous, but I could be wrong.

Thanks,
Steve

ANSWER: Dear Steve, This is a very interesting question and not one that is easy to answer.  First let me thank you for your patience as I have had a crashed hard drive and replacement since you wrote, and I have also consulted with another expert about your question.
    The answer to how many plants are poisonous would be "it depends".  This may sound vague, but first let's examine what it means to be poisonous.  The plants that we think of as obviously poisonous like oleander, poison ivy, azalea, or poisonous mushrooms, make up only a small portion of plants.  In addition, the plants with medicinal properties would all be considered poisonous, if they were used in excessive amounts.  There are plants that we don't consider poisonous at all, like apple, but the seeds can be toxic.  The fig is edible, but the sap is poisonous as a skin irritant.  Tomatoes and potatoes are edible, although many members of the Solanum family are poisonous, but tomato and potato plants are poisonous, and green potatoes are toxic.  Bermuda grass is poisonous at some times in the year to cattle.  Chocolate is delicious to us, but poisonous to dogs.  Holly berries are toxic to us, but the birds love them.  Pokeweed is edible when young, but becomes poisonous when it has berries.  Tapioca comes from a poisonous plant.  So you see, this question goes on and on, because there are many varying degrees of poisonous, and many different ways that a plant may be poisonous.  According to the sampler, 1 out of 15 would be only 6.6%, but I think that a more accurate guess would be that perhaps 30% of plants may be considered poisonous at some time.  This is a very fascinating topic, and I appreciate you bringing it up.  Please write back if you have further questions, or any more thoughts on this subject.  This would make a fantastic thesis paper!  Melissa

---------- FOLLOW-UP ----------

QUESTION: Hi Melissa,
I understand that what one considers poisonous would "depend". So then my questions would be:
1) What are the chances that out of 15 New England plants, all of them have traditional medicinal uses?
and
2) What are the chances that out of 15 such plants plants, only one of them is *deadly* to human beings? (Given that all plants, and just about everything else, can be toxic to someone, or everyone if the dose is too high?) Put another way, when I looked up the medicinal uses of these 15 plants, only one plant had a specific on it that it was poisonous, even though it had some medicinal uses. The other 14 only listed medicinal uses, with no such toxicity label.

What are my chances of randomly picking 15 New England plants, and only finding that listed in the description on one of them?

I'm trying to get a bead on whether my result is unusual or about what you'd expect in any random sampling of 15 plants, in other words.

Thanks,
Steve

ANSWER: Dear Steve,  1.  Out of 15 New England plants, I think the chances that all of them would have medicinal uses would be very slim.  Probably the average would be about 7 to 10 would have medicinal uses.  2.  Out of 15 such plants, 1 of 15 being deadly would probably be about right.  I would expect 1 to 3 on average out of 15 would be deadly.  So the deadly plant would be about typical but the medicinal ones would not be.  However, one more thing to consider in this fascinating question, would be they typical plant choices for a sampler.  I actually have a bit of background in this subject as well, and generally most plant samplers have three basic themes: herbs (mostly cooking), wild flowers, and medicinal. The medicinal samplers are pretty rare, and often there would be a few plants with only dye properties, or something that was an "old wives" remedy. Now even in the cooking herb and wild flowers you would usually find that about half would have medicinal uses, and that is generally because almost every plant has several uses, some we may know about and others that we don't.  Anyway, I hope this clarifies the question.  I actually did a lot of extra research on samplers because I was curious.  Melissa

---------- FOLLOW-UP ----------

QUESTION: I understand what you're saying, but it's not really what I'm asking.

What I was defining as "poisonous" was: when you look up a plant online for its medicinal qualities, and the description says that it is poisonous (and yet still had some traditional medicinal uses in the 19th Century).

So now we have operationally defined "poisonous" as I was defining it. Given that definition, my question is, what are the statistical odds of randomly choosing 15 plants native to New England, and only one of them having that description?

Thanks
Stephen

Answer
Dear Steve, I cannot say exactly what the statistical odds would be since as fascinating as this subject is, to my knowledge and research, no one has actually done this type of experiment.  However, with the information I have accumulated on this subject, I think the chances of randomly choosing 15 plants native to New England, the odds that only one will be defined to be poisonous are probably about 10-20%.  The chances that 2 would be poisonous would be about 40-50%, the same for 3 (40-50%) and for 4 would be back down to 10-20%.  Therefore the bell curve would probably place the general average at 2-3 per 15 being poisonous and somewhat lower for only 1 plant being poisonous, around 10-20%.  I will be looking for studies that are conducted on this subject, as I find it a very intriguing question trying to find how many plants in a given population will be poisonous. Melissa