Innovation can take its sweet time

 kw: book reviews, nonfiction, biology, culture, sociology, evolution, innovation

The cover of the book really catches the eye, being decorated with lovely shells of the Cuban land snail Polymita picta. Interestingly, these snails are never mentioned in the text. These golf-ball-sized snails are considered by many to be the most beautiful snails in the world.

For your delectation, here are some more, from one lot in Berlin's Museum für Naturkunde. 

I happen to consider a different species the "most beautiful", and I'll show some examples below.

The term Sleeping Beauties is used by practitioners of scientometrics (measuring the effectiveness of scientific ideas), to describe articles or monographs that receive few or no citations by other scientists, for years or decades, until they suddenly become "popular", possibly receiving hundreds or thousands of citations in literature. A real "blockbuster" idea might then be cited, and even enter the popular press, for many years. One learns this rather late in the book Sleeping Beauties: The Mystery of Dormant Innovations in Nature and Culture by Andreas Wagner.

The thesis of the book is that innovations crop up all the time but are seldom "on time". They may need to arise a few times before factors in the environment permit them to spread. The first example in the book is grass. Grass evolved about 70 million years ago, but widespread grasslands did not appear until about 25 million years ago in South America and Asia, and 10-15 million years ago in North America. From that time, grasses radiated into thousands of species, from cm-high Alpine grasses to 20-meter-high timber bamboo. Although grasses had several innovations that made them more efficient—growth from the base rather than the tip; drought and heat resistance; SiO2 granules in the tissues—these did not favor them when Earth was mostly moist and tropical. Only in certain isolated areas did the environment favor grasses over other plants. Then the environment changed, and so did the kind of plant cover found on the major continents. The evolution of C4 photosynthesis, which makes plants that use it well adapted to CO2 levels below 1000 ppm, and a change of climate to warmer and drier across large swathes of continents, favored grasses even more over shrubs and trees. Today, even though many decry the rising CO2 level—it is more than 400 ppm now—trees and shrubs are struggling and grow more slowly than they did while dinosaurs were around, while grasses do quite well.

Similar cases are found throughout the natural world. About one-third of the book explains what has happened in evolutionary, cellular, and molecular terms. Basically, every biological entity undergoes many small DNA mutations in various cells yearly. Some are damaging and cause the cell to die. Some affect how a cell grows and it becomes cancerous, which can cause the entire animal to die a few weeks or months later. Most seem to be neutral, in that they don't change the protein a gene makes or the way a regulatory sequence works. Finally, some improve the lot of that cell; as it divides, whatever it is doing (making digestive enzymes, making muscle move, releasing a hormone) might be done a bit better so that daughter cells attain that advantage. In a multicellular animal or plant (a metazoan) a mutation must arise in a germ cell, one that will become an egg or sperm, for the trait to enter the next generation. 

The author argues that small mutations such as SNPs (Single Nucleotide Polymorphisms) occur with great frequency. Some mutations can affect larger things. A SNP in just the right place can disable a regulatory sequence, so that a key protein is no longer made. Sometimes an error in DNA copying causes a whole gene to be duplicated. Then if you have copy A and copy B of that gene, and copy B is the next one to incur a SNP or other mutation that changes its function (usually decreasing it), the organism can get along fine as long as copy A works well. Then further changes to copy B might occur until it is sufficiently modified that it might take on another function. However, let us be clear that this doesn't happen all in one organism, but over several generations. In prokaryotes (bacteria and archaea), where the entire organism is a single cell, it is easier to see how this can lead to rapid evolution. In eukaryotes (protozoa and all metazoa, including almost everything big enough to see), a mutation, beneficial or not, must pass through the bottleneck of reproduction, occurring in a germ cell, or it will not affect the next generation. I think this is not made clear enough in the text. The author deals almost exclusively with single-celled organisms.

I'll step aside from the book's content for a moment. The statistics of metazoan reproduction are such a steep hill for new mutations to climb that it would seem that their evolution must almost come to a complete halt. But there is a "third sex" that transmits genetic information horizontally between organisms, including across species boundaries: viruses, retroviruses in particular. Although some retroviruses (such as HIV) cause disease, others do not. Let's call one of these apparently harmless viruses Friendly Human Virus, or FHV. You'd never notice an infection by FHV, it is so low-level. For a period of time the infected cells would produce FHV particles, and in the process they would incorporate the FHV genome into the cell's DNA. Not all infected cells die. About 8% of human DNA consists of hundreds of "endogenous retroviruses", stored more-or-less intact in our "noncoding DNA" (it used to be called "junk DNA"). Some of the FHV particles your body sheds will include copies of some of your own DNA that the new virus particles picked up during their manufacture, which is a rather sloppy process. When someone is infected with FHV that came from you, cells in that person's body get some of your DNA incorporated into their genomes. The "Viral 8%" is part of a "library" of potentially useful genes that may get activated by other mutations later on. Contrast this 8% with the amount of your own DNA that is actually used to make proteins (2%) and regulatory sequences including ribozymes (another 2%).

There is a lot of terminology in the chapters that form the "meat" of the author's argument. It's OK, it's all well described and defined. By the middle of the book a reader can understand how certain kinds of innovation in evolutionary traits occur over and over, until other factors in the environment one change or another something favorable, something that will spread in a population.

The second half of the book discusses similar concepts as they relate to culture. Culture is not only a human thing, although until a generation or two ago, everyone thought only humans had culture. Consider certain small British birds with bills that can puncture the heavy foil lid of a milk bottle. For many years the milkman left milk bottles on people's porches once or twice weekly. At some point, a bird managed to cut its way into a milk bottle to get a beak-full of milk. Birds observe one another, and when another bird saw the feat, it tried it. Soon, the practice spread throughout the British Isles. Milk is now dispensed in a different way, which thwarts the birds. This is an example that seems to have spread rapidly, but the details are interesting. A little records-gathering has shown that the practice was confined to a small area for a few years, and then spread rapidly after that delay. I haven't learned if it happened more than once, but the author of Beauties posits that every time an innovation is researched in detail, it is found to have occurred several or many times, but only the most recent is usually remembered. He discusses the use of citrus fruits to combat scurvy: it was discovered several times during a period of two centuries before finally "taking hold" in the British Navy, to the point that British sailors are still called Limeys.

In general (but see below), the book is a great pleasure to read. There is a lot to think about, a lot to learn. Evolution has more going on than I'd have thought.

As promised, here is an example of what I consider the most beautiful species of snail, the Splendid Jewel Snail, Liguus fasciatus splendidus. The next picture has a few examples of other subspecies of jewel snails.

These are found primarily in Florida, on the trees and shrubs of "hammocks" (the local pronunciation of "hummocks", elevated, shrubby places in the Everglades).

I have a quibble. On page 181 the author discusses writing systems that were studied in detail by Mark Changizi and colleagues, including Latin, Arabic, Hebrew, Greek, Cyrillic, Chinese and Cherokee. They sorted the characters in a script according to the number of strokes. For example, "O" and "I" have one stroke (when written without serifs), while "M" and "W" have four. Analyzed this way, there are great similarities among writing systems, and he states, "Most characters can be written with three strokes, some with one or two, but none has more than four, regardless of the writing system." I think Dr. Wagner missed something. This is ridiculous if one includes the logographic scripts, primarily Chinese, Japanese and Hangul (Korean). The image at right is scanned from the spine of my Japanese wife's "big dictionary". It reads "Complete Kanji Dictionary Ocean", where "ocean" is an emphatic modifier of "complete"; the corpus deals with more than 80,000 logographs. From top down, the number of strokes are 17, 13, 13, and 10. If one wishes to separately treat the portions that don't touch, dividing a single logogram into its root and "the rest", many of those still have more than four strokes each. Hangul characters usually have fewer strokes than Chinese, but seldom fewer than 5 or 6. I dug around among other languages of Asia. A "swirly" script like Tamil may have letters with few strokes in the formal sense, but some single strokes swirl around so much they may as well be three or four, to which another stroke or two are added. I don't know how one even counts strokes in Sanskrit! And then there's classical Mayan. The characters are very complex, and quite variable depending on the writer. Dozens of strokes are typical. A key concept when looking at a Mayan stela: the carved glyphs are intended to appear a though they were written with a brush in a paper codex (of which few survive).  Probably 95% of the world's scripts follow the "4-or-less" tendency, but it is not universal.

Secondly, as to the structure of the book. The endnotes occupy 38 pages, and more than half the space is devoted to long notes that could have been incorporated into the main body of text, which is 240 pages. Such asides take off from the main text in a way that it's often necessary to go back to the original page and re-read what came before the superscript, to get the thread of the discussion. Bad practice. Dear authors, don't put "asides" in endnotes, just keep them in the main text or leave them out. If something is worth saying, that's where to say it. Reserve endnotes for citations or very (very!) brief elucidations.