This is a compilation of my "evidence for evolution" posts to talk.origins. It is not in any particular order because each post can stand on its own.
I have edited some portions of these posts to compress the file.
I would like to thank David Paschall-Zimbal for mailing me a copy of #1 (I had not saved it).
- Chris Colby
|A look at cichlid fish in Lake Victoria||Human evolution|
|Introduction to speciation theory||Double endosymbionts|
|Isolation mediated by microorganisms||Some data from the rockhunters|
|The basics of sexual selection||Corn, caterpillars and wasps|
|Experimental reversal of parental investment||Observed speciation|
|Whale with legs||Exons and "The Curly Shuffle"|
|Introduction to polymerase chain reaction (PCR)||The universe of exons|
|DNA sequenced from 17-20 MY old magnolia||Introns of ancient origins|
|Sexual selection 2||Sex ratios|
|Trait correlates with preference||Crossbills|
|Sperm competition in 13 lined ground squirrels||Mimics: a classic textbook example trashed|
|How do squirrels "know" 3.8 is the magic number?||For female isopods, size doesn't matter|
|Swordfish and female preference||The red wolf Canis rufus|
|Genetic drift and Mullers ratchet||Directed mutagenesis|
|Mullers ratchet in an RNA virus||Mutations when needed?|
A look at cichlid fish in Lake Victoria
This is the first in what I hope to be a series of postings. In the series I hope to accomplish two things, establish that evolution is an active branch of mainstream science and that there is indeed an overwhelming amount of evidence in favor of the idea of evolution. Note that no single post is meant to be a proof, just another piece of evidence that supports the theory of evolution.
In the October 11th (1990) issue of Nature, Meyer et.al. present of paper aimed at establishing if the cichlid fish species of Lake Victoria (Africa) are monophyletic or polyphyletic. (If they all share a recent common ancestor in that lake or came from separate lineages that invaded the lake). In their paper they sequenced a 363 bp part of the cytochrome b gene and a 440 bp segment of mitochondrial DNA from what is called the control region. They sequenced these genes from several species of fish in the lake and a few species from relatively nearby lakes.
What they found was the sequences in the Lake Victoria species of fish were all very similar, but they were different from the sequences of fish in nearby lakes. All the sequences are listed in the paper.
They came to the conclusion that this indicated the cichlid species of Lake Victoria all derive from a recent common ancestor in the lake. They estimate the time of divergence at about 200,000 years ago based on a model that assumes mutations are relatively constant over time. (The lake, incidentally, had been independently dated to be 250,000 - 275,000 years old)
The News and Views section of that issue has an overview of the paper written by John Avise. Also, the cover photo of this issue consists of a picture of several of these fish.
Meyer, et. al., 1990, Monophyletic origin of Lake Victoria cichlid fishes suggested by mitochondrial DNA sequences, Nature 347: 550-553
Introduction to speciation theory
This is part two in my series of postings on studies of an evolutionary nature. As I said in part one, I have two goals in for this series. One, to show that evolution is an accepted branch of mainstream science. And two, that contrary to the continual assertions of creationists, there is an overwhelming amount of data in favor of the theory of evolution. Again, note that no single post is intended as a stand alone proof. This post is divided into two section, an introduction (the part you are reading) to provide a bit of background, and the actual summary of the paper discussed.
Speciation occurs when two (or more possibly) subsets of a formerly interbreeding population become reproductively isolated. For many years, speciation theorists thought that virtually all speciation occured when the two subsets of the population where separated by geographical boundaries. (ie, the species became split by a river, mountain range or a small group migrated out of the main region inhabited by the species.) Reproductive isolation followed physical isolation as the two, now separate lineages, diverged. This could occur for many reasons, for example mating rituals grew different or chromosome numbers changed etc. etc. In any case the end result would be that the two lineages could no longer interbreed if they encountered each other. (Incidentally this type of speciation is called allopatric speciation).
A second type of speciation, sympatric speciation, occurs when two lineages of a formerly interbreeding population diverge to the point of reproductive isolation while still residing in the same locale. This was first demonstrated to occur by Guy Bush working on the Apple maggot fly Rhagoletis pomenella.
The paper I will outline here is one found in the August 9, 1990 issue of Nature. I will continue this discussion in my next post.
Isolation mediated by microorganisms
In the paper outlined here (Breeuwer and Werren, 1990) the authors examine two species of wasps living sympatrically (in the same area).
Wasps (like ants, bees and termites) are haplodiploid organisms. In these organisms, females develop from fertilized eggs (so there is a male and a female contribution to the genome (i.e. sperm and egg)) while males develop from unfertilized eggs (so there is no male contribution to the male genome).
The authors of the paper experimented with two species of wasps, N. vitripennis and N. giraluti. They noticed that when they crossed individuals from different species, only males were produced. In other words, fertilization was not occuring. They found out that this was the result of microorganisms in the cytoplasm of the gametes destroying the males chromosomes from his sperm.
Microorganisms had been seen in the cytoplasm of the eggs of these species, but this alone did not prove that they were the cause of the reproductive isolation. So what they did was feed some wasps food that contained tetracycline, which kills microorganisms, and cross the wasps again. What they found was, in crosses in which all the microorganisms had been killed, the two species produced both male and female offspring. In crosses where the parents gametes still harbored the microorganisms, only males were produced in interspecific crosses. (Note that intraspecific crosses (matings in the same species) always produced male and female offspring)) Therefore they concluded that the microorganisms made it unable for the sperm from a different species of wasp to fertilize the females egg. This worked bidirectionally (N. vitripennis females to N. giraluti males and N. giraluti females to N. vitripennis males). The microorganisms did not, however, inhibit males and females of the same species from producing offspring of both sexes.
The authors then went on to speculate that microorganism induced reproductive isolation may be a quick way for sympatric speciation to occur. The paper also list some other cases of similar events occuring in other organisms.
Breeuwer and Werren, 1990, Microorganisms associated with chromosome destruction and reproductive isolation between two insect species, Nature 346: 558 - 560
The basics of sexual selection
This is part three in my series of postings. In this post I describe a paper presented in the July 12, 1990 issue of Nature dealing with sexual selection in katydids (an insect). I am going to break this up into two articles, one to outline the underlying theory and another to describe the experiment.
The paper I will outline deals with sexual selection. It is well accepted that the most intense competition an organism faces is with members of its own species. Many species tend to have limited diets and habitat requirements, and an organism must compete with members of his own species to secure these necessities. Of primary importance, however, is procuring a mate. If an organism fails to do that it's genes are eliminated from the gene pool. (Note that in nature there is never enough food, habitat and/or mates to go around. There are always more offspring produced in a population than will be able to reproduce.)
In many (if not most) animal systems, females choose the males they wish to mate with. Conversely, males compete for access to females. For example many male birds defend a territory in order to attract females. In many mammals (ie sheep) the males (rams) engage in contests to determine which male gets to mate. Obviously the female will choose the male who wins because her sons will then have the genes for winning these contests and females will choose them. [as a sidenote this kind of "logic" on the part of females can lead to what is called "runaway sexual selection". This occurs when the traits favored by sexual selection become linked with the genes for preference of that trait. This can often push the system in such a way that traits with a lower survival value are favored because their sexual attractiveness outweighs their negative survival value. The tail of the male peacock is an oft-cited example of this - but that's another story].
But why should females be the one's who choose? Why don't females compete for access to males? To answer this question, Darwin speculated that the sex that contributed more energy to the production of the offspring would be the sex that would be able to exercise preference. His theory of sexual selection was later expanded upon by Williams and Trivers.
In most animal systems it is clearly the female who devotes the most energy to the production of offspring. The female gamete (egg) is many times larger than the male gamete (sperm). In addition, in mammals, females must carry the offspring until birth. And furthermore females of many species provide the lions share of parental investment after the offspring has been born.
In the paper I will present in the next article the authors experimentally test the hypothesis that the sex devoting the most energy to the production of offspring will be the sex that exerts a choice amongst mates.
Experimental reversal of parental investment
In their paper, the authors (Gwynne and Simmons, 1990) experiment on a katydid of, as yet, unnamed species and genus. This species of katydid was observed to be highly variable in male contribution to parental investment. In these insects, the males transfers a spermatophore to the female after copulation. The spermatophore contains the ampulla, which contains the sperm, and the spermatophylax, which the female eats. The spermatophylax has been demonstrated to increase both the number and fitness of offspring sired by the male (it is a source of nutrition to the female).
In their experiment the authors set up two cages. In cage one (the control) the katydids were allowed to feed on the pollen of their host plants. In cage two the katydids were allowed to feed on the pollen, but were also provided a nutritional supplement (the experimental cage). Therefore, in the control cage (with limited food) the value of the males spermatophore is much greater to the female. Females were introduced to both cages and their behavior was observed.
In the control cage (with limited food) the males exerted a mating preference and females competed for mating opportunities with males. This is because, with a scarcity of food, the male spermatophore became a valuable asset.
In the experimental cage, the females exerted the mating preference because with an abundance of food, the male spermatophore was not such a valuable asset. In this way the authors showed that (in katydids at least) the parental investment is the determining factor in courtship roles (i.e. which sex exerts the mating preference)
Gwynne and Simmons, 1990, Experimental reversal of courtship roles in an insect, Nature 346: 172 - 174
Whale with legs
This is my fourth posting in my "evidence for evolution" series. This will be a short one. It's a short, gee-whiz paper from Science. In my next post (tomorrow, maybe) I'll explain a paper in Nature in which the authors sequenced DNA from a 17-20 MY old magnolia leaf. I'll tell what they found (it's cool) and how they did it (also cool).
In the July 13, 1990 issue of science, Gingerich et. al. report on an interesting fossil found in Egypt. It is a whale with feet. The skeleton is of the species Basilosaurus isis. This whale lived in the Eocene period (in Egypt (then under water obviously)).
Current cetacea (whales), as you are no doubt aware, do not have external hind limbs. But whales, which are mammals, evolved from terrestrial mammals. This fossil, therefore, is a link between the two. The skeleton they show is long (16 m) and serpentine. The authors believe this whale hunted in shallow mangrove or seagrass habitat. It's hind limb has a short femur and a slightly shorter fibula and tibia. It has no thumb and a greatly reduced second digit. The other three fingers are quite long (relatively). In short, another variation of the basic mammalian leg.
The authors speculate that the limbs were tucked in close to the body while the whale was swimming (and the topography of the bones suggest that they are correct). Furthermore, they go on to speculate that the limbs served as a copulatory guide for the whale.
The one thing I didn't like about the paper was a lack of actual photographs of the specimen. They gave graphs and schematic diagrams of all the salient features, but no photos. I would think that in a paper of this nature, a picture would have been worth a thousand words. Maybe they are working on the reconstruction and want to complete it before display.
Gingerich, et. el., 1990, Hind Limbs of Eocene Basilosaurus: Evidence of Feet in Whales, Science 249: 154-156
Introduction to polymerase chain reaction (PCR)
This is my fifth posting in my "evidence for evolution" series. In this post I will explain a paper in the April 12, 1990 issue of Nature in which the authors sequence a 17-20 million (yes, thats million) year old DNA sequence from the chloroplast of a fossilized Magnolia plant. I will use this post to make two points (besides the usual). One, to explain the significance of their actual results. And two, to introduce you to a new molecular biological technique that has opened up a vast horizon of possible molecular evolutionary studies. The technique is called polymerase chain reaction (or PCR for short). This first article describes the technique. The second article will describe it's application. This article assumes some knowledge of basic molecular biology. I give a reference for a more detailed discussion near the end.
PCR is a technique that allows a researcher to pick a region of DNA from a very small sample and amplify it to some usable quantity. It works by iterating cycles in which only the region of interest is amplified.
At the beginning of a cycle the DNA is double stranded (I'll call the strands the + and - strands). The DNA is then heated and the strands come apart. Then the DNA is cooled. As it cools, primers bind the DNA. These primers are short oligonucleotides chosen by the experimenter and added to the DNA mixture at the beginning. They flank the region to be amplified. One binds to the + strand and the other binds to the - strand. Their 3' ends both face the region to be amplified (remember DNA is synthesized in the 5' to 3' direction) so that polymerization can only occur in that region. A DNA polymerase then begins adding nucleotides to the 3' end of both primers, synthesizing a new - and + strand of the region of interest. Next, the reaction mix, (which includes the DNA sample, the primers, single nucleotides and the polymerase) is again heated and then cooled. This is repeated many times.
The result is the following. In the first cycle the + and - strand serve as a template and a new - and + (respectively) copy of the area of interest is made. When the cycle is repeated the primers now have more sites to bind to, the original sample DNA sites and the newly synthesized DNA sites. As the cycles continue, the number of possible primer binding sites doubles each time. Therefore in a short amount of time a negligible amount of DNA can be amplified to a workable quantity. This is because the amount of templates is geometrically increasing each cycle.
This is extremely hard to portray in words. A diagram of this technique makes things crystal clear. Many biologists I know, including myself, when first exposed to the idea of PCR said, "Why the hell didn't I think of that?". It is a very powerful and elegant technique. For a good, accessable overview (with the pictures to ram the idea home) see the April, 1990 issue of Scientific American (p 56, The Unusual Origin of the Polymerase Chain Reaction).
One further thing is worth mentioning. When you heat the DNA, everything else in the reaction mix is going to be heated along with it. At the temperature DNA denatures (strands separate) proteins from most organisms (like DNA polymerases) also come apart. This presents a problem. Either the researcher would have to add new polymerase each cycle, or a heat stable polymerase would have to be found. In fact, a heat stable polymerase has been found and is used for PCR. The polymerase is called Taq polymerase. It is call Taq because it comes from the organism Thermus aquaticus, a bacteria that lives in thermal vents in the ocean. Since the organism lives in water averaging close to boiling, it's DNA polymerase is stable at these high temps. And, therefore the Taq polymerase can be added to the reaction mix at the beginning and will remain active throughout all the cycles.
DNA sequenced from 17-20 MY old magnolia
In the paper I explain here, the authors (Golenberg et. al., 1990) sequenced an 820 bp region (the rbcL gene) from the chloroplast DNA of a compression fossil of a magnolia.
[A brief explanation of chloroplasts (and their DNA): Chloroplasts are organelles found in the cells of plants. They are the site of photosynthesis. These organelles are autonomously replicating (i.e. there replication is not tied to the cell cycle.) They contain their own genome, a single, circular "chromosome". DNA sequences of their "genomes" and their autonomous nature led Lynn Margulis to speculate that chloroplasts were once free living organisms that later became endosymbionts in other cells. She also thinks this explains the presence of mitochondrian in cells (as well as basal bodies). This is now generally accepted. But, that's another story]
The fossil leaf they extracted the DNA was from a compression fossil formed when the leaf sank to the bottom of a lake. The conditions were very anoxic (lacking in oxygen) and as a result the fossil was in very good condition. In the News and Views section of the same journal they show a photo of the fossil; the leaf was still green! And, as you will see, it still contained DNA. They authors mention that many well preserved compression fossils were recovered. These fossils were from organisms living in the Miocene, 17 - 20 million years ago!
Anyway, the authors extracted what DNA they could from the fossil and amplified the rbcL gene via PCR. The primers they used were 30 bp oligonucleotides synthesized to match the sequence of Zea Mays (corn). Since rbcL codes for a necessary protein, ribulose-1,5-bisphosphate carboxylase, they expected the sequence to be conserved enough for the primers to bind. It was. They also ran some tests to insure the sequence they got was actually from the fossil and not an outside contaminant. It was.
The sequence of the fossil and two extant species of magnolia are given along with one other plant species. The fossil magnolia, given the species name Magnolia latahensis, yielded a sequence similar, but distinct from the extant species of magnolia. The magnolia sequences (fossil and extant) formed a cluster distinct from sequences of closely related species (tulips and petunias for example).
The authors conclude that the sequence they got was from the fossil and that the fossil was from a now extinct species of Magnolia.
The power of this technique (PCR) suggests many applications for evolutionary biologists. Any organism in which the tissue is intact can potentially yield enough DNA to sequence. (This includes insects in amber, wooly mammoths and museum specimens) This knowledge can be used to resolve phylogenies of extinct organisms. Also, if enough samples are available, one could estimate the genetic diversity of past populations of organisms and how it changed through time. There has already been a paper of this nature in Journal of Molecular Evolution. In that paper the researcher traced the genetic diversity of Kangaroo Rats of California. Someone in my lab is doing the same thing on an endangered species of beetle here in Massachusetts. She is getting the DNA from pinned museum specimens that go back over one hundred years.
Golenberg, et.al., 1990, Chloroplast DNA sequence from a Miocene Magnolia species, Nature 344: 656 - 658
Sexual selection 2
In this post I present two models of sexual selection and a paper that tests one of the predictions of both models. The first article in the post will be an exposition of the theory and the second article will be a discussion of the paper.
Darwin, and others, noticed that in many species males developed prominent secondary sexual characteristics. A few oft cited examples are the peacocks tail, coloring and patterns in male birds in general, voice calls in frogs and flashes in fireflies. Many/most of these traits are a liability from the standpoint of survival, mainly because an ostentatious display to attract females is also going to catch to the eyes/ears/nose/whatever of predators. How then could natural selection favor these traits? Well, as I pointed out in a previous post, the sexual attractiveness of these traits outweighs the liability incurred for survival. A male who lives a short time, but produces many offspring is much more successful than a long lived one that produces few. His genes will eventually dominate the gene pool of his species.
There are two competing theories as to why females are attracted to male displays. One model, the "good genes" model, states that the display indicates some component of male fitness. A "good genes" advocate would say that bright coloring in male birds indicates a lack of parasites. The females are cueing on some signal, in this example color, that is correlated with some other important trait (ex. parasite load).
The second model, proposed by Fisher, is called the "runaway sexual selection" model. In his model he proposes that females develop a preference for some male trait (without regards to fitness) and then mate with these males. The offspring of these matings will therefore have the genes for both the trait and the preference for the trait. Note, these genes would be expressed in the males and females respectively. As a result the process snowballs out of control until natural selection brings it into check. An example to clarify.
Suppose, due to some quirk of brain chemistry, female birds of one species prefer males with longer than average tail feathers. Males in the population with longer than average feather will therefore produce more offspring than the short feathered males. So in the next generation, the average tail feather length will increase. As the generations progress, tail feather length will increase becuase females prefer not a specific length tail, but tails a little longer than average. Eventually tail feather length will increase to the point were the liability to survival is matched by the sexual attractiveness of the trait and an equilibrium will be established. Note that in many exotic birds male plumage is often very showy and many species do in fact have males with greatly elongated feathers. In some cases these feathers are shed after the breeding season.
In both of these models, which are not mutually exclusive, it is predicted that female mating preference will be correlated with the male trait. In the first case because the trait is a signal for some other, underlying beneficial trait. In the second case because the the genes for the trait and preference for the trait are, or become linked.
In the paper I will present, the authors test this prediction. Their paper is not an attempt to discriminate between these two models. If the common prediction of both of these models turned out to be false, then both the models would have to be given the boot. That is the justification for the study.
Trait correlates with preference
In the paper I discuss here, the authors (Houde and Endler,1990) conduct experiments on the guppy Poecilia reticulata. They collected these fish from 7 different streams that harbor these species. Each stream differed in the color pattern of male fish residing there. Male guppies had orange coloring covering from between 5% to 17% of their body, depending on which stream they came from.
They experimented by placing 6 males and 6 females in a tank and measuring the sexual attractiveness of the males. This was calculated as percentage of male displays that elicited a response from the female. In each separate experiment all the males were from one locale and all the females were from the same or another locale. They tested most, but not all, of the possible combinations of male/females.
They found that, female guppies from streams where males had large amounts of orange coloring strongly prefered male guppies with large amounts of orange to males with less orange. In populations where males had low amounts of orange coloring the females had no real preference with respect to coloring. The preference exhibited by females in the first sentence was, of course, statistically significant.
They interpreted this as, in the populations where coloring is prominent, evolution of female preference is correlated with the evolution of the male trait. In the populations where coloring is less prominent, there is no association between the male trait and the female preference.
The authors also mention a few factors that may confuse the issue. It had previously been shown that females in lightly predated waters favored brightly colored males more than females in heavily predated water.
In addition, a similar experiment by Kodrick and Brown had shown that females always prefered prominently colored males. They point out however, that these fish were from highly inbred lab stocks whereas Houde and Endler used fish recently sampled from nature (all the fish were less than three generations removed from the wild).
To conclude, the authors reach the conclusion that female preference and male trait are correlated in populations where the male trait is prominent. This was a prediction of both the "good genes" and the "runaway sexual selection" model.
Houde and Endler, 1990, Correlated Evolution of Female Mating Preferences and Male Color Pattern in the Guppy Poecilia reticulata, Science 248: 1405 - 1408
Sperm competition in 13 lined ground squirrels
Here's number seven in my series. It's about sperm competition and male mate choice in 13 lined ground squirrels. As I have said before, each post is just the summary of some current paper published in a mainstream peer reviewed journal. This shows that evolutionary biology is a valid. productive branch of science and is recognized as such by the scientific community as a whole. No article is meant as a capsule proof of evolution.
In most species, females choose the males they wish to mate with. This is not the case in the thirteen lined ground squirrel, Spermophilus tridecemlineatus. In this system, oestrous females mate with any male that approaches them. On average a female will have two matings. The first male to mate will sire more of the offspring than the second (this is due to sperm competition). The ratio of first male offspring to second male offspring is modulated by two factors: delay between matings and duration of second mating. The longer the delay between the first and second mating, the less offspring the second male will sire. He can increase this number, however, by increasing copulatory time.
So, when a male arrives at a female who is already being courted he has two choices. (note, the first male on the scene is always the first to mate) He can wait until the first male leaves, or attempt to find a new female (hopefully an unmated one). As it turns out, females are scarce enough that it usually pays for the second male to wait. Siring fewer offspring is preferable to not finding a mate and siring none. However, males had been observed in the field rejecting certain females (ones who had mated awhile earlier) and searching for a new mate rather than going for the sure copulation.
The authors worked out a mathematical model (a fairly simple one) that showed, after a long enough time has passed since the first mating, the second male is going to sire a negligible amount of the female litter (due to sperm competition, remember the first male sires more and the proportion gets larger as time goes on). In this case the probability (although low) of producing offspring from an unmated female that he still has to go locate is greater than the probability of producing offspring from the female he has located. (Actually it's a bit more complicated than this, but this simplifies the picture without (IMHO) distorting it) The author calculated that the critical time to be 3.8 hours, after that a male should reject a previously mated female.
The authors then observed the squirrels mating and observed that second matings did, in fact, decrease in time. They also found that, on average, males would reject a previously mated female if she had mated 3.82 hours earlier.
The authors concluded that, since the behavior of squirrels closely matched their predictions. And, since their predictions were formulated based on sperm competition; sperm competition is most likely the factor determining male acceptance/rejection of mated females in 13 lined ground squirrels.
Schwagmeyer and Parker, 1990, Male mate choice as predicted by sperm competition in thirteen lined ground squirrels, Nature 348: 62 - 64
How do squirrels "know" 3.8 is the magic number?
In regards to my previous squirrel post (actually two), the thought just crossed my mind that some people might get the wrong idea (or heaven forbid want to ridicule evolution by making a straw man of what I said) about how male ground squirrels "know" to reject previously mated females.
First off I would like to make it quite clear that the squirrels do not need to be trained in math to determine this. They don't avoid previously mated females after 3.8 hours because they understand the underlying mathematical model, but because natural selection favors males who "know" 3.8 is the magic number. Allow me to elaborate.
If a male happens upon a female who had mated, oh lets say 2.4 hours previously, decides to go looking for a new mate, (on average) he would sire less offspring than if he would have waited. Likewise, if a male waits 5 hours after the first mating for his chance, he will (on average) produce less offspring than had he wandered off to search for a new mate. However, males who, for whatever reason, go searching for mates after 3.8 hours will on average produce more offspring than males who wait any other amount of time. And as time goes on their offspring (who "know" to start searching after 3.8 hours) will come to make up a larger and larger percentage of the gene pool. Natural selection will favor males who search for a new mate when the female they find has mated 3.8 hours or more ago.
So males don't need to run around with calculators to figure out how long to wait, the answer has been passed on to them by their male ancestors who, by chance, hit upon the right length of time.
One last question could be asked. How do males know if and how long ago the female mated? I don't know the answer to this. Any thirteen lined squirrel experts out there? It could be any number of things. Even a rough estimate could be beneficial to the male.
Swordfish and female preference
In one of my earlier posts in this series, I presented two (non mutually exclusive) models of sexual selection. Those were the "good genes" model and the "runaway sexual selection" model. Well, there is actually a third model out there also (which does not exclude the others). I'm not aware of any name for it, I'll just call it the "existing female preference" model. According to this model, females have a built in preference for a certain type of male, even if that type of male does not exist.
The paper I summarize here is in the Nov 9, 1990 issue of Science. In the article, the author claims that, in swordfish, the female preference for males with swords existed before males had swords.
Within the genus Xiphophorus there are swordless platyfish and swordtails. The swordless state is considered to be ancestral. Basolo (the author) experimented with females of the species X. maculatus. Males of this species are swordless. He placed a female in the center of an aquarium that was sectioned into three areas. On one side, he placed a normal male. On the other he placed a male with an artificial sword attached. She noted that the female prefered (stayed on that side of tank and offered mating displays) the male with the artificial sword. The experiment was redone and males switched sides (to control for side bias). The result was the same, the female prefered the male with the sword to the swordless male.
The author further experimented to determine if it was the sword itself the female was cueing on. To do this she repeated the above experiment except in this case both males had artificial swords. One sword was colored, the other was opaque (clear plastic). In this case the female prefered the male with the colored sword. The control (for side preference) was also run. In addition, the author removed the swords and switched them between males and ran the tests (and controls) again. The results were once again, the same. The female prefered the male with the visible sword.
So, the data she collected were. [small aside, yes the word "data" is plural. "Datum" is the singular. Computer types simply misused the term often enough that it has become accepted in computer literature]
Basolo, 1990, Female preference predates the evolution of the sword in swordtail fish, Science 250: 808 - 810
Genetic drift and Mullers ratchet
Here's a switch, I'll try posting something productive instead of flaming people. I'll discuss here a paper in a recent issue of Nature. The author, Lin Chao, examined the RNA virus phi 6 to see if Muller ratchet was operating. I'll post this in two parts. In part one I'll explain what Mullers ratchet and genetic drift is. In part two I'll summarize the paper and explain it's significance.
H.J. Muller proposed, in 1964, one reason why sex may be beneficial to organisms. In a strictly asexual lineage, recombination is not possible (in sexual lineages it, of course, is). Thus, any mutation that occurs in an asexual lineage can only be corrected in one of two manners. The back mutation can occur or a compensating mutation can occur. Since mutations occur at random, the probability that the next mutation occuring in the lineage is the back mutation is low. Thus, each new mutation the lineage absorbs is likely to be a unique mutation. And, since mutations are most often deleterious; an asexual lineage is expected to decrease continually in fitness. Compensating mutations are also highly unlikely. This continual decrease in fitness, driven by mutations, is called Mullers ratchet. The term comes from the idea that each mutation moves the "ratchet" one notch forward and it cannot be moved back.
Sexual lineages have one other option to overcome mutations, recombination. If a gene is mutated in a sexual organism, recombination can occur with it's mate's homologous gene. Thus the offspring will have a nonmutated gene. If a sexual population has several different mutations in various genes in it's gene pool; it is possible through recombination to reconstruct an unmutated progeny. Recombination is several orders of magnitude more common than mutation, so it can easily "take care of" mutations as they arise. Some (most?) biologists think this is why sex evolved (and continues to this day). It eliminates the operation of Mullers ratchet (because organisms can shuffle all the "good genes" in the gene pool into one organism).
In order to understand the paper I will outline in my 2nd post, one must understand one more concept, genetic drift. I'll only explain this briefly.
Genetic drift is caused by a binomial sampling error of the gene pool. In a finite population (as all biological populations are) the gametes contributing to the next generation are a sample of the alleles in the gene pool. As anyone who has any grasp of statistics can tell you; the smaller a sample, the less likely you are to get an accurate description of the population. So, in populations that undergo a bottleneck (a severe reduction in numbers), the sample of alleles going to the next generation is a small sample of the population gene pool. Thus, the frequency of each allele in the following generation will be different in the next generation due solely to chance (binomial sampling error to be specific). [Note: this is assuming natural selection is not operating on the allele in question. Natural selection also changes allele frequencies.] The greater the bottleneck, the more severe the sampling error, or genetic drift, is. [Drift occurs to some degree in all population whether they are bottlenecked or not. The smaller the population, the greater effect drift as.]
Drift relates to Mullers ratchet in the following manner. When a mutation occurs in an asexual lineage, only one organism has the mutation. The rest of the organisms are unmutated. If the mutation is only slightly deleterious, it can increase via drift and eventually the unmutated version of the gene can be lossed. When this occurs, the ratchet has clicked a notch and can't be reversed. (The unmutated gene is lost from the population barring a back or compensatory mutation) Of course, to strictly asexual lineages, there is no such thing as a population. Each organism is it's own species. But, there are precious few strictly asexual organisms in the world. Most asexual lineages find some way to "mix and match" genes with those like them, and (as Deaddog could attest) those not really all that much like them. So, in that case, the population of organisms is meaningful.
Mullers ratchet in an RNA virus
In this paper Lin Chao propagated 20 lineages of the RNA virus, phi 6. This virus was chosen for two reasons. One, it is asexual. (Actually, it has three distinct regions that can be recombined, but recombination can not occur within these regions.) And two, it has a mutation rate several orders of magnitude higher than similar DNA viruses. (In addition, DNA viruses reproduce sexually.) All 20 lineages derived from a single parent virus.
In each lineage he subjected the virus to 40 growth cycles. Each cycle consisted of picking a single virus and growing up a population of 8*10^9 viruses from it. So, the virus was subjected to 40 bottlenecks to intensify drift. If the single virus chosen contained a mutation, the mutation could not be rectified. The ratchet had clicked a notch. (Intensifying drift corrected for the fact that a small amount of recombination is possible in this virus as I mentioned before.)
At the end of the forty cycles he measured the fitness of each of the 20 lineages (compared to the original parent virus). (Fitness of each lineage was measured three times.) He found that each of the 20 lineages differed markedly in fitness. One lineage increased in fitness by 6%, all others decreased in fitness. One lineage decreased to 28% of the parents fitness. The average of the lineages was 78% as fit as the parent (the 95% confidence interval did not include 1 (fitness of parent virus)).
The author concluded that the (highly significant/ P=0.0001) decrease in fitness was due to Mullers ratchet. Each lineage continued to absorb mutations it could not repair. Of course the 6% increase in fitness was an interesting result. No real satisfying explanation of that was given. (If Mullers ratchet was assumed to be operating in the past, however, one possibility immediately springs to mind.)
The paper is (IMHO) important because Mullers ratchet looks good on paper, but it had only been demonstrated once before (in ciliates. Incidentally, allowing them then to have sex stopped the ratchet.). Given that it is one of major reasons sex is thought to have evolved, it's nice to have some empirical evidence that the phenomena actually exists.
Chao, 1990, Fitness of an RNA virus decreased by Muller's ratchet, Nature 348: 454 - 455
Larry and I recently had a flamewar, er... scientific discussion about evolution in humans. I just saw a paper concerning human evolution in PNAS (Proceedings of the National Academy of Sciences) and thought I would summarize it. This bears only tangentially on that discussion.
The authors of this paper (a bunch of people from Cavilli- Sforza's lab) set out to draw a phylogeny of 5 human populations and determine whether the differences in the populations were due to natural selection or genetic drift. They gathered data on 100 genetic polymorphisms from people from these 5 groups: two groups of African pygmies, Europeans, Chinese and Melanesians. A polymorphism is a trait (in this case a gene sequence) that is variable in a population. For example, eye color in humans is a polymorphisms.
Phylogenies are drawn by comparing gene sequences and assuming that sequences more similar to each other are more closely related than sequences less similar to each other. [For a brief intro into the theory behind this see Li and Graur, 1991, Fundamentals of Molecular Evolution, Sinauer. There's a little more to it than I'm letting on. However, phylogeny construction is (IMHO) so unbearably boring I don't want to get into the details here.] They arrive at a tree that shows the African populations branching off from the others about 100,000 years ago. (Estimating time of divergence assumes a constant rate of mutation - the relationship of the sequences do not. IMHO, it is not a great idea to automatically assume mutation rates are constant.) Next the Melanesian stock split off from the European/Chinese lineage. Then the Europeans and Chinese split. Finally (in the other half of the tree) the two African stocks separated.
This tree, however, has serious problems. I won't get into them but basically there are a series of checks you can run to see if the tree the computer spits out is reasonable. This tree wasn't. For one thing the tree required Europeans to have an incredibly slow rate of evolution compared to the other populations. The authors find this unlikely although they add (are you out there Larry?) that the population explosion due to the agricultural revolution may have frozen drift and slowed evolution in Europeans by 20-25%.
Using some historical evidence the authors make the assumption that the European stock was an admixture of two other lineages. They then feed the numbers back into the computer and get the following tree. The first split is again the African/others bifurcation. Next the Chinese and the Melanesians split off. Then the European population is formed as a hybrid of the Chinese and as yet undifferentiated African stock. Finally the two African stocks diverge. The authors conclude this tree is more reasonable.
Next they tried to determine if the distribution of polymorphisms is due to drift or selection. They did this by calculating the Fst value for each polymorphism. Fst values are a measure of the variation in a subpopulation with respect to variation in the pooled population. (For details about Fst see Hartl and Clark, 1989, Principles of Population Genetics, Sinaeur.) They determined a distribution of Fst based solely on a model of drift and compared that to the numbers they calculated. They rejected the null (P=0.0023). There were too many high and low Fst values (and not enough in the middle, therefore) to be consistent with drift alone. Extraordinarily high values of Fst indicate disruptive selection. Very low values indicate stabilizing selection.
So basically the authors constructed a phylogeny of 5 human groups they felt was reasonable and determined that some of the differences in the gene pools of these groups was due to natural selection. I thought the paper was pretty good although sketchy in some portions. In any case, a reasonable preliminary data set and interpretation.
Bowcock, et. al., 1991, Drift, admixture and selection in human evolution: A study with DNA polymorphisms. PNAS 88: 893-843
Chloroplasts and mitochondria are organelles within eukaryotic cells (cells of organisms other than bacteria, which do not have organelles). These organelles have their own genetic material. It has been shown previously that organellar DNA is much more similar to bacteria than to nuclear DNA from eukaryotes. This, and other evidence, led scientists to the now widely held belief that these organelles were once free-living prokaryotic cells that began living in proto-eukaryotic cells and eventually the two types of cells required each others presence for existence. They were obligate endosymbionts. It's worth noting that organelles still reproduce autonomously within eukaryotic cells.
Recently, a paper in Nature provided evidence for a double endosymbiotic event in cryptomonad algae. Several lines of evidence led researchers to conclude this double event had taken place. First, most chloroplasts are double-membraned (one membrane from the protoeukaryotic cell, one from the endosymbiont bacteria). Chloroplasts from cryptomonad algae have more than two membranes. Also, these chloroplasts contain what is called a nucleomorph, a DNA containing structure thought to be the vestige of a eukaryotic nucleus. (Prokaryotes and organelles don't have a membrane bound nucleus, their DNA just "floats free".)
The clincher came when the researchers amplified up regions of the 18S rRNA gene (using PCR). They found two different length sequences that they called Nu and Nm. Nu they believe to be from the nuclear DNA of the algae and Nm from the nucleomorph (they are still trying to get rigorous proof of this.) The two sequences were very divergent. The Nu was similar to nuclear DNA from amoeboid protozoans and the Nm sequence is similar to red algae. The authors conclude that cryptomonad algae is a chimera of two endosymbiotic events. First a endosymbiotic event in which red algae was formed, then this eukaryotic red algae being taken into a protozoan creating the crytomonad algae.
Douglas, et. al., 1991, Cryptomonad algae are evolutionary chimaeras of two phylogenetically distinct unicellular eukaryotes, Nature 350: 148-150
Some data from the rockhunters
Hey everyone! Here's post number nine in my series. In this one (which should be short) I summarize a couple of paleontological papers. Since I'm not a rockhunter myself, I won't give too much detail.
The first paper is a report from Science by Jeram, et. al. In it they describe fossils of land animals from the Silurian. They found an arachnid (spider) and two centipedes. The kicker of the paper was that land was not supposed to be colonized by animals by the Silurian. But, finding predatory arthropods indicates a stable ecosystem containing animals much sooner than expected. [Aside to the good guys; isn't it nice to have a theory that is enriched, rather than embarrassed, by new data] This finding even made it's way to the popular press, my mom sent me a newspaper clipping (probably KC Star - I can't remember) about it.
The second paper appeared in Nature and was authored by Pilbeam et. al. In this paper they describe two recently found Sivapithecus humeri and they discuss the hypothesis that it was closely related to the genus Pongo. The upshot of the paper is, previous skull specimens of Sivapithecus indicated that it was probably closely related to Pongo, however, the newly found humeri are not at all similar to Pongo The authors conclude that the data is not sufficient to make a decision at this time.
If you want the full, gory, jargon-laden description, plus the photos of these fossils check out the refs. There is also an article about evolution of arthropods in that same issue of Science, but it's not really that interesting (to me at least).
Jeram, et. al., 1990, Land Animals in the Silurian: Arachnids and Myriapods from Shropshire, England, Science 250: 658 - 660
Pilbeam, et. al., 1990, New Sivapithecus humeri from Pakistan and the relationship of Sivapithecus and Pongo, Nature 348: 237 - 238
Corn, caterpillars and wasps
Well, I haven't heard any creationists on this board claim recently that there is no evidence for evolution, but I'll keep this series going since all the mail I've got concerning it has been favorable. I'll summarize here a paper from the most recent issue of Science.
In this paper, Turlings, et. al. investigate the interactions of corn plants, caterpillars and parasitic wasps. The wasps parasitize the caterpillars that, in turn, eat corn. The authors found that corn, when eaten by caterpillars, releases chemicals (terpenoid volatiles) that attracts wasps.
To determine what stimulus caused the release of these chemicals Turlings tested the following leaf types with respect to their ability to attract wasps: 1) leaves that caterpillars had eaten 2) leaves that were mechanically damaged (cut w/ razor blade) 3) leaves with caterpillar saliva on them. Note that the first type of leaf would have both mechanical damage and caterpillar saliva on it. It had been previously established that wasps were attracted by terpenoids.
The authors found that the first type of leaf (caterpillar chewed) attracted the most wasps. They concluded that a combination of damage and saliva were required to efficiently attract wasps.
In addition to measuring wasp attraction, they analyzed the chemicals released by the corn by gas chromotography. This was to insure that terpenoids were indeed being released. They were, so Turlings concluded it was the terpenoids that was attracting the wasps (and these terpenoids were only produced in response to caterpillar damage).
It had previously been shown that plants produce chemicals to ward off grazers. Most of these chemicals, however, work in a straightforward fashion. Bug eats chemical; bug dies. This is one of the first papers to demonstrate a chemical defense that works in a more roundabout way. Bug eats plant. Plant releases wasp attracting chemical. Wasp eats bug.
The authors do not discount the possibility that the terpenoids may also harm the caterpillars in some direct way. But, the primary value of the terpenoids to the corn is its ability to attract a predator of the caterpillar. There is more to the paper, but I just wanted to hit the highlights.
In the recent Nature there are a couple of very interesting articles (about wrens). I'll try to get around to summarizing them this week sometime.
If you are wondering what this has to do with evolution ask yourself this question, how did this system arrive at this point? Remember Steve Timm claimed that creationists (some? most? all?) believe that before "the fall" there were no predators.
It is easy to construct a plausible way for this system to reach the point it is now at given evolutionary theory. I don't see how you can given a creationist scenario. Both the corn and wasps must change in the interim between the supposed fall and present time. Creationism provides no mechanism for change.
Turlings, et. al., Exploitation of Herbivore-Induced Plant Odors by Host-Seeking Parasitic Wasps, Science 250: 1251 - 1252
Kathleen Hunt([email protected]) writes:
6) "Though evolution has been studied for years, scientists have never observed a single species evolving". So what? Evolution has been studied for just over a hundred years. Speciation takes *LONGER* than just over a hundred years. If you just study evolution for about a hundred years, all you would expect to see is microevolution within species, and perhaps the splitting off of subspecies who might be on the road to speciation. Scientists *have* observed both of these events.Creationists seem to want to define species evolving solely in terms of speciation. Microevolutionary change doesn't seem to fit their bill as evolution. In fact I just responded via email today to some guy who didn't understand how the English moths had anything to do with evolution. (To be fair, I'm not sure if he was a creationist or just didn't get my point.) As Kathleen pointed out, evolution has been observed (microevolutionary changes and the beginnings of speciation). Most creationists (as well as many evolutionists, perhaps) would be surprised to know that speciation has been observed!
In the genus Tragopogon (a plant genus consisting mostly of diploids), two new species (T. mirus and T. miscellus) have evolved. This occured within the past 50-60 years. The new species are allopolyploid descendents of two separate diploid parent species.
Here's how it happened. The new species were formed when one diploid species fertilized a different diploid species and produced a tetraploid offspring. This tetraploid offspring could not fertilize or be fertilized by either of it's two parent species types. It is reproductively isolated, the definition of a species (well, the most common definition, at least.) The paper I have corresponding to this are great. One new species, T. mirus has arisen at least three separate times.
So, speciation has been observed in case they bring up that again. In fact, it happened instantly in this case. Plants are amazingly plastic in regards to genetics, so it really isn't all that surprizing that the first (as far as I know) observed speciation event would be something like this in plants.
Soltis and Soltis, 1989, [the title is mangled on my photocopy], Amer. J. Bot. 76(8): 1119 - 1124
Roose and Gottlieb, 1976, Genetic and Biochemical Consequences of Polyploidy in Tragopogon, Evolution 30: 818 - 830
(The Soltis paper looks at chloroplast DNA, the Roose paper examines allozyme data. Both are, IMHO, fairly decent papers.)
Exons and "The Curly Shuffle"
I've mentioned the term exon shuffling in several of my posts, so I might as well get around to explaining what the hell I'm talking about. This is especially true since there is a paper in this weeks Science about the "Exon Universe" that will be the second part of this article. In this introduction to the paper, I'll explain a little about gene structure and what exon shuffling is. (Keep in mind that DNA codes for RNA which codes for protein)
The bacteria E. Coli was used in most of the very first molecular genetics experiments. When the first genes were sequenced from it, it was found that all the information for the protein lay in one continuous stretch (an open reading frame (ORF)). It came as a bit of a surprise when the first eukaryotic genes were sequenced and this was not the case. It seemed typical eukaryotic genes contained several open reading frames of DNA interrupted by sequences of DNA that did not code for anything. The coding regions were dubbed exons and the intervening sequences were dubbed introns.
It was soon found out that exons commonly coded for a functional domain or subunit of a protein. In other words, that introns often separated useful "building blocks" of proteins. Of course this led to speculation that, perhaps exons could be duplicated, deleted or "mixed and matched" from an existing gene to create a whole new gene with a new function. If a whole gene was duplicated for instance (this is fairly common), one gene could continue doing its job while the other is free to evolve a new function by swapping exons with other duplicated genes. This is what exon shuffling refers to.
Of course, this would be a great way to gain new useful genes and their corresponding proteins in a hurry. And, it wasn't too long before exon shuffling was confirmed to have happened. Two genes, low density lipoprotein (LDL) receptor gene and the epidermal growth factor (EGF) were shown to be mosaic genes. Although they were functionally unrelated, they shared a few common exons.
It may seem a bit farfetched to those who don't know much about molecular genetics that exons could just whiz all over the genome and conveniently plunk down in a useful place. In fact there are plenty of mechanisms for moving DNA from one part of the genome to another. I'll mention a couple.
One is gene conversion. This is a phenomena by which one stretch of DNA "erases" another stretch and copies itself in it's place. The mechanism is well known, but I don't have time to explain it. Any molecular bio text will have that info.
Another is transposition. This is when a stretch of DNA simply excises itself from one part of genome and moves itself to another. Transposons are pieces of DNA that do this. Many biologists (including myself) tend to think of them as molecular parasites. They don't do any good to the cell or organism. But since they move around the genome so much, it's hard to get rid of them. Transposons carry a few genes with them, usually only the genes required for their own movement.
If two transposons surround a stretch of DNA, they can carry that stretch of DNA along next time they both move if they move as one big unit.
These processes aren't directed or cognizant in any way, so an exon doesn't know to get shuffled to the right place. In fact, often an exon (or transposon) will plop down in the middle of a functional gene. The result is one dead organism. But, occasionally a good rearrangement will take place. It's a hit or miss phenomena.
So, theres an explanation of exon shuffling and a bit of info as to how it could happen. Tomorrow, I'll try to post a summary of the paper in Science.
The universe of exons
O.K., here's a summary of the paper. It's not that long really, but very dense. I'll summarize the high points and just warn you that I'm leaving out some stuff.
The paper is called "How Big is the Universe of Exons". Recall that an exon typically encodes on functional domain of a protein (for example a DNA binding domain). Duplicate genes can "swap" introns and quickly evolve new proteins. A homologous DNA binding exon, for example, might be found in many entirely unrelated gene, indicating it was imported intact from another gene. This "prefab" construction of genes is called exon shuffling.
The authors of the paper made the following assumption in the beginning of the study. Since introns (the sequences that intersperse between exons) are found in all eukaryote taxa and they typically are in the same place in homologous proteins, the intron/exon structure of genes must be ancestral. The competing point of view is that introns are rather new and spread through all taxa as transposon-like elements. Some intron placement lends credence to this view, but, IMHO, most introns were probably present in the progeonote (latest common ancestor to all living organisms). Some introns invaded later. As an aside I will mention that bacteria do not contain introns, some biologists take this to mean that they "dropped" their introns to streamline their genomes. Others take this as proof that introns invaded after the divergence of prokaryotes and eukaryotes. For what it's worth, I favor the first hypothesis.
The authors then set out to calculate how many exons would it take to account for all the proteins we have in all organisms today. This assumes modern day proteins did not each evolve slowly but were assembled by throwing together domains until something worked.
To do this they plugged their computer into the Genbank and EMBL databases and basically looked at every gene ever sequenced (a bit of an exaggeration). They then went through and narrowed the list of sequences down to non-homologous genes and non-homologous sequences within genes. For example, if the alcohol dehydrogenase sequence from one species was used, the sequences from other species were thrown out. Likewise, if a gene had more than one domain that was identical (not uncommon) the "extra" domains were deleted. All this was done in an attempt to eliminate duplicate exons from known homologous sources. (Note: all sequences were first "transcribed" from DNA sequences to amino acid sequences via the universal genetic code - this was done by computer)
Much mathematical/statistical/computer simulation mayhem followed 8-) I'll supply the reference for those who want to wade through the gory details. (I'm still mulling over some of the analysis) Basically, however, what they did can be explained as followed. From the sample of genes they took out of the database, they made pairwise comparisons and checked how many identical exons they had. They used this sample number as an estimate of the total of identical exons in the population (all organisms). They concluded that between 1000 and 7000 exons were needed to create all the proteins we see today. A rather small number, all things considered. (Boy, don't you love hand waving. I think I almost broke my wrist 8-) At least now I understand the appeal of creationism ;-) )
At the end of the paper a considerable amount of time is spent examining all the possible assumptions and consequent errors that could be included in the study. They are rather numerous, but the authors do their best to deal with them. They range from the chances of forming two identical exons by chance to homologous exons diverging in amino acid sequence, but not function. Some problems would cause the estimate of total exons too small, others would cause the number to be too large.
Well, it certainly was an interesting paper; I'll give them that. And, I would guess that they are probably not off by too many orders of magnitude 8-)
Dorit et. al., 1990, How Big is the Universe of Exons, Science 250: 1377 - 1382
Introns of ancient origins
This is number 13 in my series of postings about current research in evolution. I'll summarize two papers from a recent issue of Science, both of which basically reported the same finding. I'm kind of pressed for time today, so this will be a bare bones summary. But, as always, I'll supply the references.
First a bit of background. In eukaryotes, (basically all organisms except bacteria) genes typically are not found as a single uninterrupted reading frame. There are sequences interspersed within the coding region of genes. They are excised after the DNA is translated into RNA. These excised DNA sequences are called introns (the coding DNA sequences are called exons).
In the two papers I will summarize, the authors present evidence of an ancient origin for introns.
According to the endosymbiotic hypothesis of eukaryote evolution, modern day chloroplasts are the descendents of ancient cyanobacteria. These cyanobacteria were engulfed by an ancient cell and a symbiotic relationship was established such that the cyanobacteria simply continued to live inside the engulfing cell. There are also free living cyanobacteria alive today. The authors of the papers document the presence of an intron in a gene of both modern day cyanobacteria and chloroplasts. In both cases this intron is the same type in all the genes looked at (it is a group I intron) and it is also in the same position. They argue that this implies the intron was present in the gene before ancient cyanobacteria split into its two present day lineages (modern cyanobacteria and chloroplasts).
In the first paper the authors document a group I intron in the same position in the leucine tRNA gene in two species of Anabena (cyanobacteria) and in the chloroplasts of several land plants (bean, liverwort, maize, rice and tobacco). In the second paper the authors (a different bunch of fellows) show a group I intron in the leucine tRNA gene in five species of cyanobacteria and many chloroplasts from very different plants.
From these data the authors (in both papers) argue that this is evidence for the intron predating the split of modern cyanobacteria and chloroplasts. If the common ancestor of these two groups (ancient cyanobacteria) had this intron in that position, it's current distribution can be explained by simple inheritance; both lineages retained it. The alternate explanation would be that the intron invaded all these lineages. Group I introns are mobile in some lineages; they can excise themselves from one stretch of DNA and insert themselves in another. However, it is highly unlikely that the same type of intron would plunk down in the same spot in all these genes. The first hypothesis (the intron was in the common ancestor) is, IMHO, much more likely.
Xu, et. al., Bacterial Origin of a Chloroplast Intron: Conserved Self-Splicing Group I Introns in Cyanobacteria, Science 250: 1566 - 1569
Kuhsel, et. al., An Ancient Group I Intron Shared by Eubacteria and Chloroplasts, Science 250: 1570 - 1572
This is part 14 in my series called "evidence for evolution". For Bob's benefit, I'll explain what this is about. In this series I post summaries of recent scientific papers about evolution. I choose the papers from mainstream, peer reviewed scientific journals (not Evolution or Journal of Molecular Evolution or any of those journals). This is to demonstrate that evolutionary biologists meet the criteria of scientific worth as judged by scientists in other fields. (As an aside, Nature publishes about one to two evolution papers per week. I have never seen a creationist paper presented there.) No single post is meant to be a capsule proof of evolution. Each is merely more evidence that it did, and does, occur. In addition, I have been chosing papers that have come out recently. This is not a compendium of classic papers, but rather stuff on the cutting edge.
I'll summarize here a paper that demonstrates evolution occuring in a laboratory situation. It appeared in a recent issue of Science .
In almost all dioecious species (species with two sexes), the sex ratio is 0.5. There are 1/2 males and 1/2 females. In most species,the Mendelian rules of inheritance explain mechanistically why this is so. For example, in humans the offspring from any one mating has a 50 percent chance of being male or female. This is because the male sperm has a 50 percent chance of containing a Y chromosome and a 50 percent chance of containing an X chromosome. Female eggs only contain X chromosomes. Individuals that are XX are female, individuals that are XY are male. Given any initial sex ratio, the next generations sex ratio will be 0.5. (the proof is left as an exercise to the reader) The only exception to this would be a sex ratio of 1 or 0. An all male or all female population has no hope of regaining a balanced sex ratio.
The question can be asked, is the sex ratio then just a non-adaptive consequence of the independent assortment of X and Y chromosomes in male sperm? Or, is the ratio adaptive and Mendelian assortment an adaptive trait that has evolved?
The authors of a recent paper put this to the test by studying the Atlantic silverside fish Menidia menidia. This fish has an unusual life cycle in that, during the early months of the year mostly female offspring are produced. In the summer months mostly males are produced. The bias in the sex of the offspring is induced by the water temperature. Female offspring are produced while the water is cold, males while it is warm. The sex ratio across the whole year balances out to 0.5. This sex bias is caused by temperature dependent sex determination, not temperature dependent sex mortality. In other words cold water makes baby female fish form, it doesn't kill male baby fish. The same embryo could be male or female depending on the temperature it is raised at (i.e. Mendelian segregation does not influence the sex ratio in this species.)
The authors captured hundreds of these fish and maintained them in aquaria for five to six years. Some aquaria were maintained at low temperatures, others at high temperatures. In the low temp aquaria, the populations began with mostly females. The sex ratio , for example, in one low temp tank was 0.70 (70% female) In the high temperature aquaria, the populations began with mostly males. In one of the low tanks the sex ratio was 0.18. Both of these, given the population sizes, are significantly different than 0.50.
As the experiment progressed, the sex ratios changed from the highly skewed initial conditions. In all the populations the sex ratios converged on 0.5. The trajectory of the sex ratios converging on 0.5 differed between many of the tanks. In one tank, the next and all subsequent generations were at an 0.5 sex ration. In another, it slowly converged upon 0.5. In yet another it reached 0.5, then overshot slightly, then returned. This indicates that a sex ratio of 0.5 is somehow adaptive (there is a lot of theory as to why this may be - I may bore you with it later some time) because the fish evolved from a skewed ratio to a balanced ratio. Since chromosome assortment does not determine sex in these fish (temp does), the only explanation for their convergence to 0.5 is natural selection favored fish that produced an abnormal amount of the minority sex. (If males are lacking, any fish that produces male fish will contribute more than average to the gene pool). This is a frequency-dependent kind of selection. As the sex ratio approaches 0.5, fish who produce a disproportionate amount of either sex will contribute less than average to the gene pool.
Finally, notice that evolution has occured. The experiment started with populations of fish that produced skewed sex ratios and ended with populations that produced balanced sex rations. Since the environment was held constant, the change in the populations was therefore genetic. In other words, the gene pool changed over time. This is the definition of evolution.
Of course, the authors were mainly concerned with the result of sex ratios apparently being adaptive and did not make much ado about evolution being shown to occur (for much the same reason that modern astronomers don't constantly stress, or try to prove, the earth is round). This is only one of many papers actually demonstrating evolution in the lab. There are also many demonstrating evolution occuring in the wild (any evolution text can provide these refs - or email me if you are interested). Also, as I have posted before, speciation has also been documented to occur (I'll supply these refs [2,3])
 Conover and Voorhees, 1990, Evolution of a Balanced Sex Ratio by Frequency-Dependent Selection in a Fish, Science 250: 1556 - 1558
 Roose and Gottlieb, 1976, Genetic and Biochemical Consequences of Polyploidy in Tragopogon, Evolution 30: 818 - 830
 Soltis and Soltis, 1989, [ title mangled on my photocopy 8-( ] Amer. J. Bot. 76(8): 1119 - 1124
This will be short as I am just taking a quick study break. Steve Watson kindly provided us with some info as to what the creationists are up to these days. Here's a look at what a couple of biologists have been up to recently.
A lot of "armchair evolutionary" explanations of complex traits follow the "little trait becomes a big trait" mode. In other words the complex trait begins as a barely functional abnormality and is gradually shaped by selection into a fully functional bit of morphology (or behavior or biochemistry). These "just so" stories are usually, however, completely unsupported by data (but, they aren't refuted either).
The authors in the paper I'll summarize briefly here add some weight to one "little trait becomes a big trait" explanation. Benkman and Lindholm studied the red crossbill (Loxia curvirostra) to examine how this birds strange bill evolved. The crossbill, as it's name implies, has a crossed bill. The lower bill curves to one side and the upper bill curves to the other. The unusual bill shape helps these birds extract seeds from pine cones.
Bird bills, like human toenails, can be clipped without injuring the organism. And, again like toenails, they grow back. The authors used nail clippers to trim the beaks of the birds in such a way that they were not crossed. It took about 36 days for the bills to grow back from an uncrossed state to a crossed state. The authors used this bill growth to mirror the probable phylogenetic change from uncrossed to crossbilled birds.
The authors first separated the birds into a control and an experimental group. They then measured how long it took the birds in each group to extract seeds from a cone. Both groups were statistically the same. They then clipped the beaks of the experimental group and measured, over a 36 day period, how long it took each group to removed seeds from a cone. As you would expect, the control group did not change throughout the experiment since it remained unaltered. The experimental group, however, did change. In the first day after clipping it took an average of 5.28 seconds for a bird to get at a seed. This was up from 1.34 seconds prior to clipping. As the experiment went on, the birds got better and better until at 36 days it took them only 1.68 seconds to get a seed (statistically not significantly different from 1.34). This increase in seed gathering seed was interpreted as a function of bill crossing. The authors concluded that the crossbill trait was selected every step of the way from an uncrossed ancestor, because as the bill became more and more crossed, the birds ability to quickly secure food increased. They also noted that slight bill crossings have been sighted in straight billed species of birds.
John Krebs (in the accompanying "News and Views" article) notes that the paper does not address the changes in musculature, tongue morphology and behavior that must accompany the change in bill morphology. But, he notes that these birds provide a very interesting avenue with which to pursue questions of this type.
He also notes that, according to legend, these red crossbills got their beaks crossed trying to remove the nails from Jesus Christ's cross. The red coloring of the males symbolized his blood. For some reason, I like the first explanation better 8-)
Well, there you have it. A short "gee-whiz" paper from Nature. I sort of liked it (it had a good beat and I could dance to it). Usually I don't buy these "little trait becomes big trait" arguments, but in this case at least there is a little data to back up the claim. (I should point out, before Larry accuses me of being a saltationist ;-), that I'm not implying that complex traits appear fullblown in "hopeful monsters". I just I think that it is often the case that the current utility of a trait has little or nothing to do with it's ancestral utility. Many complex traits may be exaptations, not adaptations.)
Benkman and Lindholm, 1991, The advantages and evolution of a morphological novelty, Nature 349: 519-521
Mimics: a classic textbook example trashed
Ted, in his own charming way, has explained that all science is bogus because it is based on false assumptions and that scientists are so caught up in the momentum of what they are doing, they can't go back and correct their "errors" (apparently this would involve forgetting mathematics and selectively reading old manuscripts). In any case, what gets done can't be undone (ITHO). This brings me to a recent paper in Nature. First, a little bit of ecology/evolutionary theory.
Mimicry is the phenomena of two (or more) species looking/sounding/smelling/whatever like each other. There are two types of mimicry: Batesian and Mullerian (should be an umlaut over the u).
Batesian mimicry is when one species evolves to mimic a second species that has some trait that makes it undesirable to predators. For instance, a butterfly that tastes good, but mimics a butterfly that tastes bad, may evade predation as long as bad tasting butterflies outnumber good tasting ones (it's a frequency dependent kind of thing). The palatable species benefits because it gains the reward of looking like a bad tasting species, but it doesn't pay the price; chemical toxins are costly for an animal to produce. If the palatable species becomes too numerous, the unpalatable species may suffer as predators may learn that organisms that have that pattern/coloring/sound/smell are O.K. to eat.
Mullerian mimicry is when two (or more) foul species evolve to look/sound/smell/whatever like each other. They both/all benefit because predators have only to learn one signal to discriminate species to avoid, instead of having to learn separate cues for each foul species. This is a benefit to the prey (only coincidentally a benefit to the predator) especially if the two (or more) mimetic species occur at low densities.
One of the classic example of Batesian mimicry has been the viceroy butterfly. Biology texts explain that this butterfly is a palatable mimic of two species of noxious butterflies, the monarch and the mueen. A new paper in Nature suggests that the viceroy tastes every bit as bad as monarchs and worse than queen butterflies. The authors conclude that the mimicry is a three way Mullerian mimicry instead of the viceroy being a Batesian mimic to the two Mullerian mimics, the monarch and queen.
Their experiments consisted of capturing 16 red winged blackbirds from nature and offering them butterfly abdomens and recording the response. Only abdomens were offered so that the bird could not tell species apart by subtle changes in wing color or morphology. Each bird was offered 8 viceroy, 8 monarch and 8 queen butterfly abdomens dispersed at random between 24 palatible control abdomens. The percentage of each abdomen type eaten was recorded as well as mean manipulation time and a mean response score. The response score was basically an arbitrary scale ranging from the bird ignoring the abdomen (0), through pecking once (1), partially eaten (2) to completely eaten (3). The results were as follows:
98% of the control abdomens were consumed by the birds, 68% of the queen, 41% of the viceroy and 46% of the monarch. The monarch and viceroy scores did not differ significantly; the other two classes did.
Mean manipulation time for the control abdomens was 5.3 seconds. For the other species it was: queen = 17.5 s, viceroy = 23.5 s and monarch = 31.3 s. The monarch and viceroy were again not significantly different. In addition the viceroy was not significantly different than the queen. The monarch and queens did, however, differ significantly.
Finally, the mean response score for the controls was 2.98. The queen, viceroy and monarchs scored 2.50, 1.98 and 2.10, respectively. In this case the viceroy and the monarch were the only two classes that did not differ significantly.
So, all three species were significantly less palatable than the controls. And, in two of three measures the viceroy and monarchs were (as a class) less palatable than the queen. This shows that the classic example of Batesian mimicry is actually a case of Mullerian mimicry. It also disproves Ted's notion that once science gets done, it cannot get undone.
This is my favorite kind of science paper, one in which something widely held is demonstrated to be just plain wrong.
Ritland and Brower, 1991, The viceroy butterfly is not a batesian mimic, Nature 350: 497- 498
Batesian and Mullerian mimicry involve interspecific interactions. Intraspecific (within a species) Batesian mimicry has also been documented. Monarchs obtain their toxins by sequestering cardiac glycosides of their host-plant, the milkweed. In large flocks(?) of monarchs there are many that have not spent the energy to sequester these glycosides; they are getting a "free ride". These monarchs can then invest more energy towards raising offspring than the monarchs who "play fair" and spend energy to harbor the poisons.
For female isopods, size doesn't matter
In the marine isopod Paracerceis sculpta, there are three discrete male morphologies. These are determined by a single allele change at one locus. The largest of the three males, the alpha males, defend harems of isopod females. The intermediate size male, the beta, mimics female morphology and behavior and the gamma males, the smallest of the three morphs, attempt to hide in large harems and not attract the attention of the alpha male(s). The larger the male, the slower it matures. But larger males, although they reach reproductive age later in life, live longer and therefore have more reproductively active years. In the paper I will summarize here, the authors demonstrate that each male morph enjoys equal mating success.
Male reproductive success in these isopods depends on many factors. Each male morph is able to sire roughly the same amount of offspring when isolated from other males. Differences in male reproductive success occur when males are mixed in the mating area, the spongocoel. For example if the spongocoel contains one alpha and one beta male, the beta males sires 60 percent of the offspring. If there is one alpha and a gamma male, the alpha sires 92 percent of the offspring. If there are 2 alpha males and three gamma males, each gamma males sires 33 1/3 percent of the offspring. The authors give mating success for 14 different combinations of males in the paper (all the combinations they found in nature).
They sampled isopods from a natural population for a period of two years. They found that each male morph had, on average, equal mating success and the alleles that determined male morphology were in Hardy-Weinberg equilibrium (HW equilibrium is a measure of how alleles are distributed in a population.) In the paper they present a table showing how many spongocoels were sampled with respect to each different combination of males. The table also lists how many females were in each harem. To make a long story short: the numbers of males, combinations of males and numbers of females added up such that each male morph was equally reproductively successful. Below is a summary of some of the data:
Although it appears alpha males have a higher mean # of mates, the difference is not significant (look at the standard errors in the beta and gamma males). Notice also that equal repro success does not mean equal frequency in the population; it only means that each male type is able to keep replacing itself in the population. In other words, if conditions stay the same, the ratio of alpha to beta to gamma males will stay 452:20:83.male type mean (+/- se) # of matesnumber of males --------- ------------------------ --------------- alpha1.51 (+/- 0.08)452 beta1.35 (+/- 0.44)20 gamma1.37 (+/- 0.23)83 variance within types = 3.075 variance among types = 0.003
Shuster and Wade, 1991, Equal mating success among male reproductive strategies in a marine isopod, Nature 350: 608 - 610
The red wolf Canis rufus
There has been a small amount of discussion about what is a species here on t.o (and also sci.bio) recently. A recent paper in Nature presents some interesting food for thought on this topic.
Wayne and Jenks, in a recent Nature, present a study of the mtDNA(mitochondrial DNA -- it's maternally inherited) of the endangered red wolf, Canis rufus. This species, once extending over a large range in the southeast, is now extinct in the wild. The authors examined the mtDNA sequence of red wolves (zoo animals and from DNA obtained from museum pelts from 1905 to 1930) as well as grey wolves and coyotes. (The red wolf occurred only in regions where grey wolves and coyotes were.)
When they analysed the red wolf sequences, they found that the mtDNA was either of grey wolf type or coyote type. This (along with the geographic information) lead them to conclude that the "species" red wolf is (was) actually a hybrid of the grey wolf and the coyote.
But wait, the story gets even more interesting. Grey wolves and coyotes have overlapping ranges in the northern US, but the red wolf phenotype is not present in hybrids in the north. The red wolf phenotype is not only a product of the hybridization, but of environment as well.
That's just the beginning, however, the red wolf was classified as an endangered species; but US Fish and Wildlife does not extend endangered species classification to hybrids. The authors argue, however, that the "species" former prominence in the food web -- it was the top predator in it's former range -- and the possibility that the phenotype might not be recoverable by future hybridizations -- remember, it doesn't work up north -- indicate that the red wolf deserves to remain classified as an endangered species.
Wayne and Jenks, 1991, Mitochondrial DNA analysis implying extensive hybridization of the endangered red wolf Canis rufus, Nature 351: 565 - 567
Two new papers examining the phenomena of directed mutations have recently appeared in the literature. I'll quickly review these experiments in the next post. This post is a short introduction to a few of the classic papers relevant to this issue.
In 1943, Luria and Delbruck did an experiment that led biologists to believe mutations occured at random. They started many parallel cultures of E. Coli., let them grow, then exposed them to the bacterial virus lambda. They found that the distribution of resistant cells across all the independent lines was Poisson. Some cultures had many resistant bacteria, others had few. If the phage had induced the correct mutation to occur, each independent line should have roughly the same number of mutants (a Gaussian distribution would be found across all cultures). 
Lederberg and Lederberg, in 1952, provided another experiment showing that mutations occured randomly. They grew bacteria on plates and used round pieces of felt to transfer bacteria to a replica plate. So, the pattern of colonies growing on the two plates were identical -- and the corresponding colonies on each plate all came from a single clone. Lederberg then exposed one plate to lambda and noted the colonies that survived. Then, he picked the corresponding colony on the replica plate and grew it up. All the bacteria he grew were resistant to the phage -- even though they had never been exposed to it. In other words, the mutation was present before its effect was needed. 
In 1988, Cairns questioned the experimental design of these studies. He suggested that they did indeed show that some mutations occured at random, but they did not rule out the possibility that other mutations could be directed. Lambda kills bacteria instantly, he reasoned; I'll try the experiment over with something that will slowly kill the bacteria to see if, given a chance, bacteria can direct their mutations. His paper on lactose starved bacteria suggested that some mutations were directed (i.e. the specific mutation -- and only the specific mutation -- could be induced by the cell.) However, his study drew lots of criticism becuase it left a lot of loose ends. 
In 1990, Hall released a data set that expanded on Cairn's work and met all the earlier criticisms. He showed that, under stress, some bacteria can induce a mutation to fix a "broken" gene -- and not produce (many) other mutations. In other words, the stress was not acting as generalized mutagen. The needed mutation was occuring far too often to be explained by random mutagenesis. 
The field is still divided about this topic. I'm convinced that Hall has demonstrated a class of (seemingly) non-random mutations. But, in interpreting the possible impact on evolutionary theory, one must be aware of exactly what effect has been shown and the distribution of organisms that can be affected. The only effect shown so far is a directed mutation rescuing a "dead" gene. Nobody has shown that directed mutations can create a novel phenotype.
In addition, only unicellular organisms (or organisms with totipotent cell lines) can have an evolutionary benefit from this mode of mutation. Multicellular organisms would need the directed mutation to occur in it's germ line cells for it to be evolutionarily interesting; and germ line cells are the least likely to be exposed to the stress. By the time a sperm or egg cell itself is stressed, the multicellular organism is probably dead.
None-the-less, this is a very hot area of research. A good description of exactly what is happening is being sought, as well as a mechanism to explain it. Some data indicates that there are a suite of related phenomena, for directed mutations have been claimed to: fix single base substitutions, fix frameshift mutations and correct large insertion mutations.
 Luria and Delbruck, 1943, Mutations of Bacteria from Virus Sensitivity to Virus Resistance, Genetics 28: 491 - 511
 Lederberg and Lederberg, 1952, J. Bact. 63: 399 - 406
 Cairns, et.al., 1988, The origin of mutants, Nature 335: 142 - 145
 Hall, 1990, Spontaneous Point Mutations That Occur More Often When Advantageous Than When Neutral, Genetics 126: 5 - 16
Mutations when needed?
In a recent paper in PNAS, Hall examined circumstances where a bacteria needed two independent mutations in order to survive. He concluded that, in bacteria, mutations occur more frequently when they are needed than when they are not.
He experimented on three strains of E. Coli deficient in the tryptophan (trp) operon. One strain contained a mutation at position 46 of the trpA gene. The second strain contained a mutation at position 9578 of the trpB gene and the third contained both mutations. None of the strains could produce tryptophan needed for survival or growth.
He grew up the trpA and trpB strains in media that contained trp so the mutations did not hinder their growth. Then he spread them on petri dishes that contained all their required nutrients except trp; so, only cells that mutated could survive on these petri plates. He examined the plates for evidence of growth every day for 30 days. As time went on, many revertants for each strain arose. Hall measured the reversion rate for both strains (trpA and trpB), and repeated the experiment using the trpAB strain.
Now, if the reversion mutations in trpA and trpB occured randomly, the reversion rate for trpAB should equal the reversion rate of trpA times the reversion rate of trpB. Hall found the rate to be 10^8 higher than this. (Yes, that's significant 8-).
The revertants Hall found fell into three different classes (I, II and III). Class III grew the fastest (wild type rate), Class II grew slower and class I grew the slowest. When he sequenced the genes in these revertants, he found that class III revertants produced the correct mutation such that the original amino acid was restored to trpA. Class II mutants produced a mutation such that a similar amino acid was restored to trpA -- making it functional enough to save the cell. In class I mutants, no mutation was evident in trpA -- evidently a suppressor mutation occured (a mutation in another gene, usually tRNA, to compensate for another mutation.) In the trpB region, all the classes had the correct mutation. No other mutations were found in the regions sequenced. (Hall ruled out the fact that he was selecting for bacteria with extremely high mutation rates in another part of the paper -- see the ref if you are interested.)
So, his data indicates that the bacteria could somehow induce the mutations they needed for survival -- and only those mutations.
Tomorrow, or maybe Monday, I'll summarize a paper by Cairns that demonstrates adaptive reversion mutations that involves a frameshift mutation. I'm also preparing a post about a new example of speciation.
Hall, 1991, Adaptive evolution that requires multiple spontaneous mutations: Mutations involving base substitutions, PNAS 88: 5882- 5886